Package Details: google-chrome 104.0.5112.79-1

Git Clone URL: (read-only, click to copy)
Package Base: google-chrome
Description: The popular and trusted web browser by Google (Stable Channel)
Upstream URL:
Keywords: chromium
Licenses: custom:chrome
Submitter: None
Maintainer: luzifer
Last Packager: luzifer
Votes: 2155
Popularity: 11.04
First Submitted: 2010-05-25 20:25 (UTC)
Last Updated: 2022-08-02 18:10 (UTC)

Sources (3)

Pinned Comments

luzifer commented on 2020-02-03 00:35 (UTC) (edited on 2020-08-19 11:05 (UTC) by luzifer)

When reporting this package as outdated make sure there is indeed a new version for Linux Desktop. You can have a look at the "Stable updates" tag in Release blog for this. Do not report updates for ChromeOS, Android or other platforms stable versions as updates here.

This package will automatically get updated as soon as the Debian package is available. This is checked at least once per hour.

Latest Comments

Ildar commented on 2022-07-29 08:42 (UTC)

@PC2022 I don't know if it is related but I have --password-store=basic in my ~/.config/chrome-flags.conf wo any problems in xfce + xfwm4 4.16. Passwords stored in plain text(basic mode) are not encrypted! See

PC2022 commented on 2022-07-21 23:17 (UTC)

Ever since I started using arch (and google-chrome-stable), I find that Chrome quickly forgets cookies and login sessions, so I have to re-login to many sites several times a day.

It is unrelated to upgrading Chrome, or even rebooting.

Unfortunately it's very unpredictable, sometimes it seems to lose cookies after an hour or two, and other times after a day or two. It is not due to the site setting a short cookie expiry, since it affects github, Amazon etc

A week or two ago I added --password-store=gnome to ~/.config/chrome-flags.conf, as I saw someone suggest it as a solution, but it hasn't made a difference.

Any tips on how I can fix or at least troubleshoot this? Im using awesomewm as my WM, with xfce desktop, and a few basic gnome services running.

mjr4077au commented on 2022-07-20 04:57 (UTC)

@esiqveland, I know it's been a few months now, but I had the exact issues you've described on a fresh Chrome execution where there was nothing in ~/.config.

I have composition completely disabled in /etc/X11/xorg.conf due to some issues I was having in a few games. The "fix" was to remove the line from xorg.conf, reboot, run Chrome and disable hardware acceleration, then I was able to add the line back, reboot and use Chrome as required.

If you have composition disabled in xorg.conf like I do and you have no idea why, you'd probably be best to remove the line and leave it. I only have it there for specific reasons.

dooberdoober commented on 2022-06-28 15:11 (UTC)

Replace ability to disable Side Panel Button. Its not used.

HSY commented on 2022-06-27 15:26 (UTC)

103.0.5060.53-1 Playing netflix DRM content,sometimes has problems

mattj commented on 2022-06-23 18:13 (UTC)

Anybody else's extension popup's very narrow to the extent they're unusable, with no way to make them wider?

Reezlaw commented on 2022-06-12 22:09 (UTC) (edited on 2022-06-12 22:12 (UTC) by Reezlaw)

Hi, I'm on 102.0.5005.115-1 and I have a problem saving files with "Save As", when the dialog opens and I try to change the name of the file to save, the focus shifts to the search in the same window, and then I have to click back in the name field. I type one letter and it immediately goes back to search. I have to click back to the name field for each letter I want to type for the file name. Anyone else?

P.S. typing the name in the search field, cutting and pasting in the name field works, but it's certainly not the expected behaviour. Also worth mentioning that the Save As dialog pops up unfocused

ArMor007 commented on 2022-06-05 09:44 (UTC)

Downgrading to 101.0.4951.64 (which is the latest 101 stable release) fixes the tab dragging issue.

slowbro commented on 2022-05-29 05:59 (UTC)

@rcqwerty I'm also having the tab-dragging/reordering issue, again. Looks like a bug regression in 102...

rcqwerty commented on 2022-05-26 20:55 (UTC)

Having the same problem in 102.0.5005.61-1 as we had in version 97.0.4692.71 -- as soon as I drag a tab 1 pixel it pops out into a new window, and cannot re-attach the tab.

luzifer commented on 2022-05-12 10:38 (UTC)

@JP-Ellis works fine for me using makechrootpkg as well as building in a clean docker container (my default way to build the package).

JP-Ellis commented on 2022-05-12 01:55 (UTC) (edited on 2022-05-12 02:22 (UTC) by JP-Ellis)

It appears that the PKGBUILD does not include $pkgname.install in the list of sources despite requiring it. As a result, this package does not correctly build in a chroot environment as $pkgname.install is copied into the chroot environment.

To check the error, try building the package through makechrootpkg instead of makepkg.

EDIT: I actually believe this might be an error with makechrootpkg as the PKGBUILD man page does say that .install files should automatically be added to the source files. I will report this upstream.

dvit commented on 2022-05-07 17:19 (UTC)

To enable chrome://flags/#enable-reader-mode First activate this flag chrome://flags/#temporary-unexpire-flags-m100

dtschmitz commented on 2022-05-07 17:13 (UTC)

I am running production version:

Google Chrome 101.0.4951.54 (Official Build) (64-bit) Revision 67da1aeb32cedd27634ca6634fb79cbd85d3f0ab-refs/branch-heads/4951@{#1126}

It seems that chrome://flags/#enable-reader-mode is nowhere to be found.

Please advise and thank you for all that you do!! Cheers. DTS

hexchain commented on 2022-05-02 16:15 (UTC)

What is the purpose of the "StartupWMClass" fix, and is it possible to drop it? The chromium.desktop in the official chromium package does not seem to have this line.

yu17 commented on 2022-04-12 07:14 (UTC)

I had the same issue with v100.0.4896.75 and v100.0.4896.60.

With v100.0.4896.88 everything should be fine.

Gytoo commented on 2022-04-10 14:37 (UTC) (edited on 2022-04-10 16:09 (UTC) by Gytoo)

No problem for me, everything working

>>esiqveland commented on 2022-04-10 15:12 (EEST)
>>I have had major issues with crashes and severe graphical glitches since about Chrome #>>97.x, but the latest 100.x does not even start anymore. Anyone else having issues?

esiqveland commented on 2022-04-10 12:12 (UTC)

I have had major issues with crashes and severe graphical glitches since about Chrome 97.x, but the latest 100.x does not even start anymore. Anyone else having issues?

emild commented on 2022-04-02 14:41 (UTC) (edited on 2022-04-02 14:42 (UTC) by emild)

The latest version 100.0.4896.60-1 crashes on startup. Also as per @openmindead below, cannot restore saved tabs on start and fails to restart if clicked on restore button. Don't upgrade.

[2483:2483:0402/] InitializeSandbox() called with multiple threads in process gpu-process.
[0402/153505.840407:ERROR:elf_dynamic_array_reader.h(64)] tag not found
[0402/153505.841439:ERROR:elf_dynamic_array_reader.h(64)] tag not found
Trace/breakpoint trap (core dumped

PancakeSparkle1 commented on 2022-03-31 22:07 (UTC)

Had to downgrade to version 99.0.4844.74, the latest version kept my CPU at 100% usage, with only 1 tab open, as well as not closing even with "Allow running in background" disabled, I had to manually kill it every time, using killall, already submitted a bug report to Google.

AsukaLSoryu commented on 2022-03-30 17:37 (UTC)

Getting rendering issues as of version 100.0.4896.60-1. Turning of hardware acceleration fixed it if anyone else is seeing the same issues.

openmindead commented on 2022-03-30 12:12 (UTC) (edited on 2022-04-03 14:01 (UTC) by openmindead)

The latest version 100.0.4896.60-1 cannot restore saved tabs on start and fails to restart if clicked on restore button. Be careful folks. The issue is pertinent to deb package itself (same issue in Ubuntu).

There are 2 ways to deal with it: wipe ~/.config/google-chrome and start from scratch, or just downgrade.

luzifer commented on 2022-03-02 16:10 (UTC) (edited on 2022-03-02 16:12 (UTC) by luzifer)

@yochananmarqos: The package was automatically downgraded after Google pulled 99.0.4844.51 from the stable channel of their repository… As we're just re-packaging the Debian version this package follows the most recent version available in the official repo… That one downgraded between Mar 2, 2022 1:43 PM and 2:10 PM UTC…

Current state as of 4:10 PM UTC is still v98:

Package: google-chrome-stable
Version: 98.0.4758.102-1

(Still it's indeed an interesting question why v99 was pulled from the stable channel…)

yochananmarqos commented on 2022-03-02 16:01 (UTC)

@luzifer: Why did you revert back to 98?

xunil64 commented on 2022-02-20 17:57 (UTC) (edited on 2022-02-20 18:03 (UTC) by xunil64)

WebGL is no longer working with the latest updates:

pacman -Qi google-chrom

Name : google-chrome Version : 98.0.4758.102-1

core-dump: process 12440 (chrome) of user 1026 dumped core.
Module with build-id 2ad6b24b487ac2a9c2901c8ddf457eeb465c1d5a

Module with build-id e1dcc2a88cfaafed882d09c90c668af0eed4efed

Module with build-id 2e245c2bf12f95fd8ab79b3a4be99524677cbd70

Module with build-id b710b880b0d4e9c93554fa632129db44c20632cd

Module with build-id 1f873ddb2c32ab39d0b7d8646d03680ffe99de7c

Module with build-id 4b0babfcad161c2ad0af6e59e2493258db23a331

Module with build-id 2dd6e65129a809dab828a1d26215a3f7a363fcc8

Module with build-id af8afb2331ac08c52132102ecbdc6076a8eb4c6f

Module with build-id 8eb3977b830f3012d9ead7b4def5f4d2f38ec688

full dump can be sent

cerealfordinner commented on 2022-02-08 21:00 (UTC)

Is anyone experiencing the issue where on anything after Version 94.0.4606.71 in XFCE you lose the ability to adjust the window size on the sides of the window and you can only do so from the corners?

mattj commented on 2022-02-05 22:08 (UTC)

With chrome 98.0.4758.80-1, Wayland, and the flags --enable-features=UseOzonePlatform --ozone-platform=wayland --ignore-gpu-blocklist --enable-zero-copy --enable-gpu-rasterization --enable-accelerated-video-decode --enable-features=VaapiVideoDecoder --password-store=kwallet5 the window zooms in about 4x on start after being interacted with. Anybody else getting this?

BachoSeven commented on 2022-02-05 12:58 (UTC)

Thanks for finding that out, @koshikas!

koshikas commented on 2022-02-03 06:45 (UTC) (edited on 2022-02-03 07:15 (UTC) by koshikas)

anyone else experiencing video-acceleration disabled since new chrome-98

Contents of chrome://gpu: Video Decode: Hardware accelerated

Video Acceleration Information Decoding (Legacy VideoDecodeAccelerator)
Decode h264 baseline 16x16 to 4096x4096 pixels Decode h264 main 16x16 to 4096x4096 pixels Decode h264 high 16x16 to 4096x4096 pixels

but under developer tools and intel_gpu_top no hardware acceleration is utilized.

EDIT; adding --disable-features=UseChromeOSDirectVideoDecoder got the hardware acceleration working again. now the Video Acceleration Information reads as;

Video Acceleration Information Decoding (VideoDecoder) Decode h264 baseline 16x16 to 4096x4096 pixels Decode h264 main 16x16 to 4096x4096 pixels Decode h264 high 16x16 to 4096x4096 pixels Decoding (Legacy VideoDecodeAccelerator)
Decode h264 baseline 16x16 to 4096x4096 pixels Decode h264 main 16x16 to 4096x4096 pixels Decode h264 high 16x16 to 4096x4096 pixels Encoding

babam commented on 2022-01-30 17:40 (UTC)

@ryuukk Chrome on Linux can only use HW acceleration if using an extension like NotYetAV1 or enhanced-h264ify.

ryuukk commented on 2022-01-30 02:16 (UTC)

Please update it so the process starts with:

google-chrome-stable --use-gl=desktop --enable-features=VaapiVideoDecoder

by default, so we get video decode HW accelerated by default, that should be the default that everyone should expect

otherwise people think linux is trash because no HW accell for videos, it feels bad

luzifer commented on 2022-01-29 15:09 (UTC)

@H0301ein hmm looks like the flag might does no harm (cannot test it myself) but that would be the first flag the package itself overrides from the default distribution… At moment the browser is delivered as Google packages it (with minor patches to the desktop file which I'm also hesitant to implement) while it allows users to override flags themselves using a simple config file. Looking at the chromium package the flag is set to its default value too. Therefore following this chain of arguments I'd not override that flag in the package itself. Every user can set the flag easily in their ~/.config/chrome-flags.conf and we're not sticking to an old patch when upstream changes the default value of that flag…

H0301ein commented on 2022-01-28 19:04 (UTC)

@luzifer i think it's good if add flag --ozone-platform-hint=auto to /bin/google-chrome-stable for run in both X11/wayland

alexey2baranov commented on 2022-01-26 08:29 (UTC) (edited on 2022-01-26 08:29 (UTC) by alexey2baranov)

hm. but after remove and install it becomes v97:)

alexey2baranov commented on 2022-01-26 08:26 (UTC)

As I read at January 19, 2022 The Stable channel has been updated to 97.0.4692.99 for Windows, Mac and Linux

Today in this repo I see v90.0.4430 :(

bobarch64 commented on 2022-01-20 06:00 (UTC)

@babam, maybe this is not the best place to find info on hardware acc. But better look here:

babam commented on 2022-01-15 03:41 (UTC) (edited on 2022-01-15 04:07 (UTC) by babam)

Is Chrome+Youtube able to use hardware acceleration (X11)?

In my case Chrome+Youtube can only use the hardware acceleration if I install the NotYetAv1 extension.

FFmpeg+MPV uses hardware acceleration when playing AV1 videos.


yuanhao commented on 2022-01-14 21:36 (UTC)

My youtube ad block addon is also not working properly after the update (of course and the tab drag and drop function), is it just me?

mdryden commented on 2022-01-14 14:54 (UTC) (edited on 2022-01-14 14:55 (UTC) by mdryden)

For any other Arch newbies who installed with yay and are struggling to figure out how to apply the instructions in the Wiki, perhaps I can save you a bit of time.

Assuming you've installed a prior version of chrome with yay, and haven't disabled the yay cache, yay should have cached it in ~/.cache/yay. If you do have a previously cached version, you can use the pacman command from the wiki to roll back using that path instead:

# pacman -U .cache/yay/google-chrome/google-chrome-96.0.4664.110-1-x86_64.pkg.tar.zst

archz commented on 2022-01-13 23:35 (UTC)

google-chrome-dev 99.0.4818.0-1 still has the bookmark/tab bug google-chrome 96.0.4664.110-1 works fine

akhil commented on 2022-01-13 04:47 (UTC)

I just downgraded to 96 to fix bookmarks issue. and let's wait for 99 release.

kamilst96 commented on 2022-01-12 17:05 (UTC) (edited on 2022-01-12 17:18 (UTC) by kamilst96)

All drag and drop issues fixed in version 99, currently google-chrome-dev AUR package.

Never in my life have I thought that I would need to run development version of a program to have working drag and drop. For a moment, I considered moving to Gecko or even using Windows for internet activity...

Either pull version 99, or say on version 94 as long as possible.

blackhole commented on 2022-01-12 09:11 (UTC)

Drag and drop on bookmark bar is not working. Some fix?

burz_ex commented on 2022-01-10 09:52 (UTC)

It looks like the mouse cursor is not hiding on fullscreen video playback for 97. 96 works fine.

ozmartian commented on 2022-01-09 05:45 (UTC) (edited on 2022-01-09 05:45 (UTC) by ozmartian)

FYI, while we wait for the moving tabs w/ mouse bug to be fixed, you can use keyboard shortcuts to work around the issue.


metabee commented on 2022-01-08 21:26 (UTC)


This method I shared solved my problem. Using this package didn't enable me, for example, to use Whatsapp in a different window like a different app. I'm on KDE Plasma X11.

Why it solved or the logic behind that didn't really bother me lol

Just sharing in case anyone stumbles on the same problem.

yochananmarqos commented on 2022-01-08 19:39 (UTC)

@metabee: This package uses the official DEB package, using debtap makes no sense.

There are no AUR helpers including pacaur in the Arch repos, they can't be installed with pacman. Installing one AUR helper to install another AUR helper doesn't make any sense.

metabee commented on 2022-01-08 19:15 (UTC) (edited on 2022-01-08 21:19 (UTC) by metabee)

This package doesn't allow to use chrome apps as separate apps. I instead built the .deb chrome install from the chrome official webpage using pacaur to install yay, yay to install debtap and debtap to install chrome.

sudo pacman -S pacaur

pacaur -S yay

yay -S debtap

debtap -u


debtap <<<<<<<NAME OF THE CHROME FILE.DEB>>>>>>>

sudo pacman -U <<<<<NAME OF THE FILE BUILT BY DEBPTAP>>>>>>>

now I can use chrome desktop apps

cytomax55 commented on 2022-01-08 17:35 (UTC)

Cant move tabs around and cant move around bookmarks after the last update

Fraxinas commented on 2022-01-08 10:46 (UTC)

I experienced the same problem, not being able to drag around tabs anymore with 97.0.4692.71. Rolling back to chromium-96.0.4664.93 fixed to for me, too.

Varakh commented on 2022-01-07 12:58 (UTC)

There also seem to be problems with drag and drop in latest release. Working fine with 96 version. Did not find the proper chromium bug report though.

vdwijngaert commented on 2022-01-07 10:07 (UTC)

The issue related to dragging things has been fixed in canary, it is tracked in

ozmartian commented on 2022-01-07 02:29 (UTC)

Having tab bug w/ Chrome 97 too as well as mouse cursor no longer hiding in fullscreen video mode.

GameKing505 commented on 2022-01-06 15:07 (UTC)

@leddy231 I am having the same problem with version 97.0.4692.71 - as soon as I drag a tab 1 pixel it pops out into a new window.

microgod commented on 2022-01-05 20:06 (UTC)

@leddy231 I have the same problem with moving tabs around I cant rearrange tabs or merge them, downgrading from v97.0.4692.71 to v96.0.4664.110 fixed the problem.

moonbound commented on 2022-01-05 05:02 (UTC)

@leddy231 currently experiencing the same tab bug on 97. Running EndeavourOS

leddy231 commented on 2022-01-05 02:44 (UTC)

Something broke on chrome 97 on a fresh install of Manjaro KDE, it was working fine and then the update some hour ago broke the tabs. I can no longer move tabs around, instead it pops them into a new window, and cant be merged back. Broken on google-chrome 97 and google-chrome-dev 98. Working fine on chromium 96 and google-chrome 96 (previous commit) Gonna see if I can figure out what it is and make a bug report to chromium, but just a heads up if more users report it

cerealfordinner commented on 2021-12-13 22:25 (UTC)

@Thundas, it must have broken in a later release, it was occurring on 96.0.4664.110 and 96.0.4664.93, but removing and installing .45 manually fixed it. Thank you for the help.

cerealfordinner commented on 2021-12-13 21:29 (UTC)

I'm wondering if this is affecting anyone else. Recently I have been unable to drag Chrome windows to resize them by using the left, right, or bottom edge of the window. The top-left, top-right, and top all work. I should note this is in XFCE.

Thundas commented on 2021-12-03 04:15 (UTC)

@yochananmarqos thanks, i switched to yours and its working great, its sad that the update broke angle's hardware decoding.

yochananmarqos commented on 2021-12-03 03:49 (UTC)

@Thundas: You have duplicates and old flags that do nothing. Check Command Line under chrome://gpu to see which flags are actually being used.

My flags (last two I'm testing):


Command Line:

/opt/google/chrome/google-chrome --disable-gpu-driver-bug-workarounds --enable-accelerated-video-decode --enable-crashpad --enable-features=VaapiVideoDecoder --ignore-gpu-blocklist --origin-trial-disabled-features=CaptureHandle --use-gl=desktop --enable-crashpad --flag-switches-begin --enable-gpu-rasterization --enable-zero-copy --ignore-gpu-blocklist --enable-features=VaapiVideoDecoder,CanvasOopRasterization,ChromeLabs,EnableDrDc,UnexpireFlagsM94 --flag-switches-end

Thundas commented on 2021-12-03 02:06 (UTC)

Hardware decoding works for me with the following flags on nvidia gpu 495 driver.

--ignore-gpu-blocklist --enable-gpu-rasterization --enable-zero-copy --disable-gpu-driver-bug-workarounds --enable-features=VaapiVideoDecoder --use-gl=desktop --enable-native-gpu-memory-buffers --enable-accelerated-mjpeg-decode --enable-accelerated-video --ignore-gpu-blacklist --enable-gpu-rasterization

wayneb commented on 2021-11-26 19:21 (UTC)

All maximizing issues with Chrome 96.0.4664.45 on XFCE 4.16 were fixed with an upgrade to xfwm4 v4.16.1-3. See xfwm4 4.16.1-3: handle removal of GTK_FRAME_EXTENTS

ssr commented on 2021-11-20 07:32 (UTC) (edited on 2021-11-20 08:01 (UTC) by ssr)

The link mentioned by user @luzifer doesn’t work for me, but according to my test, the latest Chromium 98.0.4716.0 has fixed this problem.

This problem can also be reproduced under google-chrome-beta 97.0.4692.20 and google-chrome-dev 98.0.4710.4.

This problem also effect YouTube, but only videos encoded as avc will have problems. Such as:

luzifer commented on 2021-11-19 19:16 (UTC) (edited on 2021-11-19 19:17 (UTC) by luzifer)

Video issues probably related to this:

Nek.12 commented on 2021-11-19 19:15 (UTC)

Confirm that on version google-chrome 96.0.4664.45-1 videos show white screen instead. Enablind raw draw did not help but rather broke things more. Disabling override software rendering list did not help

ssr commented on 2021-11-18 10:54 (UTC) (edited on 2021-11-18 11:02 (UTC) by ssr)

Same 96.0.4664.45-1 playing video issue here, Windowing system: X11 KDE 5.15.2-arch1-1

Disable hardware acceleration will working properly.

After downgrade to 95.0.4638.69-1 it working properly now.

mazlumgg commented on 2021-11-16 22:00 (UTC) (edited on 2021-11-17 00:09 (UTC) by mazlumgg)

After update 96.0.4664.45-1 version. The issue with playing video on Twitter. Windowing System: X11 GNOME Arch Linux 5.14.10-1 edit: How IDK but now it's working. I guess it was about Twitter.

wayneb commented on 2021-11-15 19:11 (UTC)

The issue with maximizing the window on XFCE (4.16) is fixed for me in 96.0.4664.45. However, the issue still exists when snapping the window to make it maximized.

alfredo.ardito commented on 2021-11-15 18:43 (UTC)

As @marcinse stated, I confirm the issue with maximizing the window on XFCE has been fixed.

marcinse commented on 2021-11-15 15:43 (UTC)

The issue with maximizing the window on XFCE (4.16) is fixed for me in 96.0.4664.45.

Kev commented on 2021-11-10 12:20 (UTC)

Starting with google-chrome 95 the user interface is becoming laggy everywhere (mouse + keyboard). This is true including at least 95.0.4638.69-1.

RIscRIpt commented on 2021-11-03 20:13 (UTC) (edited on 2021-11-03 20:14 (UTC) by RIscRIpt)

After updating to 95.0.4638.69-1 I cannot enable hardware acceleration of video decoding on X11 in any known way. Most likely that's because there's no way of disabling OzonePlatform anymore.

serhii_devrel commented on 2021-10-31 10:50 (UTC)

95.0.4638.69-1 has a problem with maximizing the window on Manjaro Linux, Xfce 4.16

aliazhar commented on 2021-10-30 16:59 (UTC)

95.0.4638.69-1 has problem with maximize window, while I downgrade to 94.0.4606.81-1 until this problem is fixed

alfredo.ardito commented on 2021-10-29 06:20 (UTC)

window maximize issue on XFCE still persists

arozbiz commented on 2021-10-28 13:04 (UTC)

I am still having the window maximizing problem on 95.0.4638.54-1, running XFCE.

michaldybczak commented on 2021-10-27 17:58 (UTC)

wayneb, I can't confirm this issue on 95.0.4638.54-1. I see that the system titlebar is finally fixed and shows correctly. Anyway, when I turn off the titlebar (CSD on), window's size is correct after maximizing (fills perfectly to screen edges on any monitor).

wayneb commented on 2021-10-27 00:03 (UTC)

Upgrading from 94.0.4606.81-1 -> 95.0.4638.54-1 caused the window to be too big (larger than the display ~16px on each edge) when maximized. Only enabling "Use system title bar and borders" allowed me to work around this issue. Presumably this has something to do with the new Ozone platform abstraction layer which is now enabled by default in Chrome 95.

somebody commented on 2021-10-22 02:04 (UTC) (edited on 2021-10-22 02:04 (UTC) by somebody)

yea this is a huge problem - i've had this problem for a long time now.
rudimentary testing (window.ondrag = (e) => console.log(e.pageX, e.pageY)) indicates that drag position is inconsistent between actual position and (0, 0)
(which matches anecdotal evidence of drags working if you wiggle your mouse in the right way)
note that:

  • iframes change the offset to (-iframe.left, or something like that
  • this also affects e.g. the bookmark bar
  • this may be touch related, which may be testable using the commandline flags for touch. not sure how to do this though.

currently on 94.0.4606.81; i think this has been an issue since around 92 (or before?) for me

kamilst96 commented on 2021-10-20 10:41 (UTC)

Version 95.0.4638.54-1 also blocks drag & drop handing by web application. I can't place Elementor widgets, and can't drag a message in Outlook to a folder.

alfredo.ardito commented on 2021-10-20 09:53 (UTC)

yep, this happens on XFCE. I'll send them a bug report ASAP. thx

alfredo.ardito commented on 2021-10-20 07:26 (UTC)

I confirm @lcp29 Version 95.0.4638.54-1 seems buggy. Maximizing exceeds screen size,

lcp29 commented on 2021-10-20 03:08 (UTC) (edited on 2021-10-20 03:08 (UTC) by lcp29)

The new 95.0.4638.54-1 version maximizes to an external padding of the screen. The previous (to 95.0.4638.54-1) 94.0.4606.81-1 version doesn't.

Noe367 commented on 2021-10-17 15:39 (UTC)

Hey, I use this chrome on Manjaro, it works flawlessly except one thing. The problem I have is in Proxmox, the kayboard layout is wrong : I use french azerty, but it uses a mix of qwerty and azerty. It only happens on Proxmox and on this version of Chrome. Is there a way to solve that ?

FabioLolix commented on 2021-10-07 20:30 (UTC)

Thanks for updating!

luzifer commented on 2021-10-05 21:06 (UTC)

@FabioLolix Thanks for the ping! I suppose I marked that read, got distracted and never re-visited that PR… I'm very sorry about that! Will take care about it this evening!

FabioLolix commented on 2021-10-05 20:45 (UTC)

Hello, do you mind adding a copy of widevine license in the default place? The PR have been open for 10 months yet is so trivial, since I made the PR you made part of it (replacing msg2 with printf), just copy the other 2 line too

leoplusclaude commented on 2021-09-30 20:23 (UTC)

If you're having trouble with "use system titlebar and borders" setting, you can just wait for patch to come. For details, please refer to

marcelroed commented on 2021-09-30 19:11 (UTC) (edited on 2021-10-01 14:37 (UTC) by marcelroed)

Since version 92/93 I've been having 5-second freezing issues on AMD iGPU Ryzen 4000. Does anyone know how to fix this? I've tried disabling all hardware acceleration but it doesn't help.

michaldybczak commented on 2021-09-24 19:44 (UTC)

Thanks @zeparu, I have the same problem with titlebar. Disabling this flag restores titlebar in Plasma, however, the whole option of turning SSD on/off is gone. It works for me, but the underlying issue is still there. The option stopped working.

zeparu commented on 2021-09-22 07:35 (UTC) (edited on 2021-09-22 07:45 (UTC) by zeparu)

just updated today, and "use system titlebar and borders" is not working anymore for me on kde

edit: disabling the flag use-ozone-platform in chrome://flags fixes it for me

debohman commented on 2021-09-10 20:24 (UTC)


jason15M71 commented on 2021-09-10 20:01 (UTC)

Are update notifications working for this package? I haven't been getting the emails.

KI9N commented on 2021-09-09 17:18 (UTC) (edited on 2021-09-09 17:19 (UTC) by KI9N)

@marcelroed I had to disable Vulkan and it's been ok since. According to chrome://gpu, Video Decode is still Hardware, but Vulkan is disabled. I don't have any flags set except --enable-features=VaapiVideoDecoder. Note: I'm using a 4th gen i5 that only has partial Vulkan support so I think that was my issue.

marcelroed commented on 2021-09-09 17:11 (UTC)

I've also been having terrible problems with lag on the latest two versions. Disabling hardware acceleration didn't help.

KI9N commented on 2021-09-05 16:06 (UTC) (edited on 2021-09-05 21:57 (UTC) by KI9N)

The latest version 93.0.4577.63-1 locks up/lags (for about 15 secs) for me all the time. I think it may have a memory leak or hardware acceleration issue. I end up getting OOM Killer messages for it in my syslogs when it happens. If I downgrade or use the beta/dev version, works fine. Be warned is all.

EDIT: The beta is also lagging so I am disabling Hardware Acceleration and retesting on the latest Stable version

Rob_H commented on 2021-09-01 23:22 (UTC)

Earlier 404 error is gone for me now. Likely a problem on Google's side as @covegeary suggested.

miguelangellv commented on 2021-09-01 17:05 (UTC)

Live captions dont work for me.

Any way to enable?

covegeary commented on 2021-09-01 14:06 (UTC)

The URL 404's for me as well, as does Seems like an issue on Google's side (

debohman commented on 2021-09-01 13:18 (UTC)

I had no trouble upgrading to 93.0.4577.63.

luzifer commented on 2021-09-01 13:14 (UTC)

Works for me at this moment:

==> Retrieving sources...
  -> Downloading google-chrome-stable_93.0.4577.63-1_amd64.deb...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 79.0M  100 79.0M    0     0  37.1M      0  0:00:02  0:00:02 --:--:-- 37.1M
  -> Found eula_text.html
  -> Found
==> Validating source files with sha512sums...
    google-chrome-stable_93.0.4577.63-1_amd64.deb ... Passed
    eula_text.html ... Passed ... Passed
==> Removing existing $srcdir/ directory...

Rob_H commented on 2021-09-01 12:53 (UTC)

Getting a 404 error trying to upgrade to the newest release.

curl: (22) The requested URL returned error: 404 ==> ERROR: Failure while downloading Aborting...

koshikas commented on 2021-08-28 05:04 (UTC)

so the latest in chrome video acceleration saga is, having merely --enable-features=VaapiVideoDecoder may not be enough. i'm on X11 and to get video acceleration functional, i need addition switch;


ozone platform is enabled by default as of late in chrome, and for some reason disables video acceleration atleast on X11.

jfabernathy commented on 2021-08-24 10:14 (UTC) (edited on 2021-08-26 13:14 (UTC) by jfabernathy)

I'm trying to use Chrome to play streaming services videos. I can do Netflix, Amazon Prime Video, but not SlingTV. I can do it on the version of Google Chrome that you get from Google directly as a .deb and install in *buntu linuxes, but not the Arch AUR version. Any ideas what's missing?

This hardware involved is Intel NUC 1165G7 Iris GFX

koshikas commented on 2021-08-07 15:09 (UTC)

@bulletmark, @kamilst96, @Nek.12, @hendry

i had my flags(set by the internal experimental flags) cease to function since 90. i could only enable hardware video acceleration by including flags in the .desktop (~/.local/share/applications/google-chrome.desktop) entry file under every Exec key;

Exec=/usr/bin/google-chrome-stable %U --enable-features=VaapiVideoDecoder

i suggest you try with your relevant flags

hendry commented on 2021-08-07 12:55 (UTC)

Video decode with vulkan on my T14s AMD GPU also doesn't work :/

Nek.12 commented on 2021-08-06 12:16 (UTC)

Have the same issue as @kamilst96. Chrome ingores flags in chrome-flags.conf. Confirmed by the page chrome://version

kamilst96 commented on 2021-08-04 18:48 (UTC)

Starting with Chrome 92, accelerated video decoding doesn't work with NVIDIA 470.57 driver - the image just smears on the screen. Anyone?

Relevant flags:



bulletmark commented on 2021-08-03 22:56 (UTC)

A trap with new version 92+ is that it does not ignore commented out options in ~/.config/chrome-flags.conf anymore.

akiirui commented on 2021-08-03 18:35 (UTC)

92.0.4515.131-1 works with systemd-resolved.

But this issue still exists:

meirg commented on 2021-08-03 14:06 (UTC)

Latest build is working now for me also with nothing downgraded or disabled

zelimir commented on 2021-08-03 06:28 (UTC)

@JoveYu Yes, it looks like it's working fine, for me anyway.

JoveYu commented on 2021-08-03 05:53 (UTC)

any one try new version 92.0.4515.131-1?

andrej commented on 2021-08-01 02:31 (UTC) (edited on 2021-08-01 02:41 (UTC) by andrej)

For some reason the current version doesn’t work at all on my desktop. No pages are loaded, just loading forever. There is no error message explaining what’s wrong. Also, all signs of hardware acceleration are switched off in chrome://gpu. Chromium works fine (and so do other browsers).

EDIT: google-chrome-beta has no such issue.

(BTW, the ideas around disabling systemd-resolved below are bad ideas. :-) Just sayin'.)

warmos commented on 2021-07-31 14:11 (UTC)

I can at least confirm that disabling systemd-resolved got me google-chrome running again

debohman commented on 2021-07-31 00:21 (UTC)

The simplest way is to just reinstall the old package with pacman. If you cd to the google-chrome repository on your system, you should see the older packages.

Just do (e.g.):

% sudo pacman -U google-chrome-91.0.4472.114-1-x86_64.pkg.tar.zst

Frnco commented on 2021-07-31 00:13 (UTC)

Downgrading is a very poor solution but does work.

1202's comment worked like a charm to me, so I made a one-liner in Bash that does it all, only asking for confirmation before removing the folder at the end, to ensure the one-liner has zero chances of deleting anything that shouldn't be deleted even if you mess with the environment variable (provided of course you use the same variable in the echo that asks for confirmation and in the rm command that actually deletes stuff). The Uppercase I in the rm is the option that prompts you for confirmation before deleting anything, while the lowercase r makes it recursive so it deletes the directory together with all its contents. The -c option in wget makes it try to continue the download if it failed previously, not strictly necessary but can help in some situations, while the -O - option means "output (-O) to stdout (-)", which pipes the contents of the file to stdout and we then pipe it into tar -xz (No need for the f option as there's no file, and the v option we usually include is just for verbosity which we also don't need. Then we cd into the folder, install using makepkg -si --boconfirm. Here -s syncs the deps and -i installs the package, while --noconfirm skips asking for confirmation, then we cd out of the folder, and finally ask the user for permission to remove the folder, which doubles as a way to make doubly sure nothing is removed by mistake.

$ hashedName=aur-710114824f61f1468346d7de4072dc041fac8177; wget -c$hashedName.tar.gz -O - | tar -xz && cd $hashedName && makepkg -si --noconfirm && cd .. && echo "Please confirm removal of folder $hashedName" && rm -rI $hashedName

If like me you use Fish Shell, setting the environment variable has to be done using Fish syntax, which is one of the reasons I included confirmation before removing anything, because when I was starting to use Fish I did run into problems when trying to run commands that I used without problem in bash but in Fish my environment variables were completely ignored and interpreted as empty strings.

Call it a trauma but yeah just to make thrice sure nobody ends up deleting anything by accident, here's a version that works in Fish Shell:

$ set hashedName aur-710114824f61f1468346d7de4072dc041fac8177; wget -c$hashedName.tar.gz -O - | tar -xz && cd $hashedName && makepkg -si --noconfirm && cd .. && echo "Please onfirm removal of folder $hashedName" && rm -rI $hashedName

If you think of a simpler way that is clearer than this and as simple to copy-paste it, please do tell.

1202 commented on 2021-07-30 09:51 (UTC)

Disabling systemd-resolved does not work for me, it messes with my other network connections. I downgraded to the latest version 91: wget then tar -zxvf aur-..., cd aur-... and makepkg -si.

b1tninja commented on 2021-07-29 18:07 (UTC)

You will have to disable systemd-resolved or downgrade to a systemd < 249 until version 93.

Messing with /etc/nssswitch.conf may be another option

redsolja commented on 2021-07-29 13:10 (UTC)

Google Chrome crashes/hangs, when using a default/vanilla Arch Linux system, with systemd-resolved enabled.

This should be fixed, as many of us use systemd-resolved.

Tyrin.price commented on 2021-07-23 02:07 (UTC)

My problem was resolved when I stopped and disabled dhcpcd.service ... that service is not needed for my setup.

nallekarhu commented on 2021-07-22 01:53 (UTC)

@koshikas Yeah, I'm not surprised, after all this is not your normal run of the mill git package.

FWIW, I have no problems with "google-chrome 92.0.4515.107-1" and systemd 249 after disabling systemd-resolved.

koshikas commented on 2021-07-22 01:07 (UTC)

the fix has been merged but will be only be available to the next build. they are not recalling the current build since, i quote;

"It seems Ubuntu is using systemd 247 in their most recent non-LTS and debian is using 247 in their testing branch. Fedora 34 is on 248 and probably won't be on 249 for a while. openSUSE tumbleweed isn't on it either. Those are our officially supported Linux distros. Definitely merge this to 93. But this is probably simple and safe enough to merge it to 92 given that it causes crashes for anyone using systemd 249."


nallekarhu commented on 2021-07-21 22:25 (UTC)

@bulletmark Hmm.. it looks to me that "chromium 92.0.4515.107-3" applies two fixes and removes the workaround in "chromium 92.0.4515.107-2", so it makes me think you don't need it.

bulletmark commented on 2021-07-21 21:59 (UTC)

@nallekarhyu, note that was the second commit applied to chromium. The first commit is required as well.

Tyrin.price commented on 2021-07-21 21:35 (UTC)

Now we just need an updated AUR package and we'll be set.

nallekarhu commented on 2021-07-21 18:44 (UTC)

Fixed on chromium side.. "chromium 92.0.4515.107-3: upstream fixes for glibc and systemd"

mappy24 commented on 2021-07-21 14:00 (UTC)

I'm not sure if this is related to the other issues being reported here, but after updating to 92.0.4515.107 this morning I find that I can no longer print as the print dialog completely crashes Chrome. Rolling back to 91.0.4472.164 has made printing work again.

When trying to print, the dialog appeared briefly with a spinning blue circle where the destination printer should have been before crashing. Running from the terminal shows this:

$ google-chrome-stable
[4616:4616:0721/] InitializeSandbox() called with multiple threads in process gpu-process.
[4571:4571:0721/] Initiated PrintPreviewHandler::HandleGetInitialSettings()
[4571:4571:0721/] Initiated PrintPreviewHandler::HandleGetPrinters()
[4571:4571:0721/] Initiated PrintPreviewHandler::HandleGetPrinterCapabilities()
Segmentation fault (core dumped)

JoveYu commented on 2021-07-21 10:40 (UTC) (edited on 2021-07-21 10:40 (UTC) by JoveYu)

if you use yay, you can use package cache

cd ~/.cache/yay/google-chrome
yay -U google-chrome-91.0.4472.164-1-x86_64.pkg.tar.zst

benoliver999 commented on 2021-07-21 10:11 (UTC)

Same issue here, also fixed by downgrading chrome.

The 'download' button on the AUR commit history page is incredibly useful.

bulletmark commented on 2021-07-21 06:15 (UTC)

@b3nn, you mean version 248.3-2, and yes that does fix it. But I suggest it is much more sensible to simply downgrade google-chrome.

b3nn commented on 2021-07-21 06:08 (UTC)

Guys, try downgrading systemd-libs to 248-5. That fixed it for me.

JayGatsby commented on 2021-07-21 03:03 (UTC)

my guess is that this only presents when using systemd-networkd as that is the only thing out of the ordinary on my system and related to networking.

JayGatsby commented on 2021-07-21 02:54 (UTC)

I'm getting this network error too. First presented in the dev channel and after the last update I'm seeing it on stable.

Tyrin.price commented on 2021-07-21 02:34 (UTC)

Chrome does nothing since update to version 92. Broken.

JoveYu commented on 2021-07-21 01:41 (UTC)

after upgrade to 92, chrome cause 100% cpu and keep rising memory after start

hendry commented on 2021-07-21 01:40 (UTC)

I'm getting "Network service crashed, restarting service" too on 92.0.4515.107-1. Now I'm wondering how to downgrade since surprisingly 91 isn't in /var/cache/pacman/pkg ... guess it's because AUR. Doh.

rodrigo21 commented on 2021-07-20 23:16 (UTC)

I'm having the same problem: Network service crashed, restarting service.

bulletmark commented on 2021-07-20 23:12 (UTC) (edited on 2021-07-20 23:13 (UTC) by bulletmark)

@yochananmarqos well it's not just us two.

yochananmarqos commented on 2021-07-20 23:03 (UTC)

@bulletmark @the-k: I can't reproduce that. There must be something you both have in common on your system causing it.

the-k commented on 2021-07-20 22:47 (UTC)

92.0.4515.107 is broken for me too, getting "Network service crashed, restarting service." Downgrading systemd didn't help, downgrading Chrome to 91.0.4472.164 did.

bulletmark commented on 2021-07-20 21:53 (UTC)

New version 92.0.4515.107-1 locks up immediately for me when starting. Gone back to 91.0.4472.164-1 which is fine.

luzifer commented on 2021-07-17 18:04 (UTC)

@noabody And the reason you would do that is which? That would break the chance for anyone to downgrade to an older version which is currently possible by just using an older PKGBUILD to build the package…

noabody commented on 2021-07-17 17:53 (UTC) (edited on 2021-07-17 17:57 (UTC) by noabody)

Apparently Google uses a single standard updated URL for the latest Debian version (beta, stable, unstable):

One could 'SKIP' the SHA check and do a blind build with version update like so:

pkgver() {
  tar -xf control.tar.xz './control' --to-command="grep -Pio '(?<=Version: )[\d\.]+'"

rjahanbakhshi commented on 2021-07-16 09:10 (UTC) (edited on 2021-07-16 09:20 (UTC) by rjahanbakhshi)

The Stable channel has been updated to 91.0.4472.164 for Windows, Mac and Linux but it's not picked up by automatic checker for this package. By manually setting pkgver to 91.0.4472.164 in the PKGBUILD the latest version is correctly installed.

yochananmarqos commented on 2021-07-16 01:04 (UTC) (edited on 2021-07-16 01:04 (UTC) by yochananmarqos)

@luzifer: Please match the optdepends() with chromium. libunity is deprecated and no longer builds, FYI.

acro5piano commented on 2021-07-15 02:51 (UTC)] Network service crashed, restarting service.

Same here at systemd 249 (249-2-arch)

Using Chromium for now

Starwalker commented on 2021-07-14 05:49 (UTC)

Google Chrome stopped working after updating systemd to 249-3, the error is:] Network service crashed, restarting service.

Seems like the issue is related to DNS resolving, as browser can still open simple pages if I use IP address.

Downgrading to systemd 248-3 resolves the issue. Meanwhile Chromium works totally fine with systemd 249-3.

NightSky commented on 2021-07-06 00:19 (UTC)

On KDE desktop environment, F11 hotkey for entering in the fullscreen mode and exiting the fullscreen mode isn't working.

Det commented on 2021-06-18 17:37 (UTC)

Man I missed google-chrome 2000 votes.

simontunnat commented on 2021-06-18 14:55 (UTC) (edited on 2021-06-18 15:05 (UTC) by simontunnat)

You are right. It works now.

luzifer commented on 2021-06-18 08:45 (UTC)

@simontunnat: That looks more like a hiccup in the AUR git repo… Just try again…

simontunnat commented on 2021-06-18 05:38 (UTC)

I get the following error message when trying to update to version "":

" error fetching google-chrome: fatal: unable to access '': Received HTTP code 502 from proxy after CONNECT "

Is there something wrong with the package definition?

Jackcode01 commented on 2021-05-30 03:35 (UTC)


Thanks for the help. Yes, --force-device-scale-factor=2 does fix the problem.

Thank you. -Jack

Parintachin commented on 2021-05-29 07:00 (UTC) (edited on 2021-05-29 07:00 (UTC) by Parintachin)


Yup; apparently it's a bug further up the chain? It came up again when Brave was updated yesterday; the guy maintaining that package was good enough to answer me.

A fix is to run Chrome with --force-device-scale-factor=2, as described here:

Jackcode01 commented on 2021-05-29 04:05 (UTC)

On Manjaro KDE with 200% scale setting, Everything is HUGE!

koshikas commented on 2021-05-26 12:47 (UTC) (edited on 2021-05-26 12:48 (UTC) by koshikas)

to get get hardware video acceleration, you need to explicitly pass; --enable-features=VaapiVideoDecoder

in the CLI. the flag “enable-accelerated-video-decode” is no more in chrome 91.

Parintachin commented on 2021-05-26 06:50 (UTC) (edited on 2021-05-28 02:52 (UTC) by Parintachin)

Using this on a high-dpi screen with scaling appears to apply the scaling twice. Everything's huge.

Manjaro KDE.

jose1711 commented on 2021-05-10 08:08 (UTC)

Hello, thank you maintaining this PKGBUILD. I wonder if a script similar to archlinux-java could be added. Currently package() results in a script called /usr/bin/google-chrome-stable but there are tools/libraries that rely on existence of google-chrome instead.

Here's an example for webbrowser python module:

python -c 'import webbrowser; webbrowser.get(); print(webbrowser._tryorder)' => ['firefox', 'xdg-open', 'epiphany', 'opera', 'links', 'elinks', 'lynx', 'w3m']

After creating a symbolic links this starts to works as expected:

ln -s /usr/bin/google-chrome-stable /usr/bin/google-chrome
python -c 'import webbrowser; webbrowser.get(); print(webbrowser._tryorder)' => ['firefox', 'xdg-open', 'epiphany', 'google-chrome', 'opera', 'links', 'elinks', 'lynx', 'w3m']

rubens.lopes commented on 2021-04-23 14:17 (UTC) (edited on 2021-04-23 14:18 (UTC) by rubens.lopes)

I'm trying to install an older version (c0a289bc52cd) but got an error and don't know how to proceed

==> Making package: google-chrome 78.0.3904.108-1 (sex 23 abr 2021 09:28:07)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
-> Found google-chrome-stable_78.0.3904.108_amd64.deb
-> Found eula_text.html
-> Found
==> Validating source files with sha512sums...
google-chrome-stable_78.0.3904.108_amd64.deb ... FAILED
eula_text.html ... Passed ... Passed
==> ERROR: One or more files did not pass the validity check!

marcosfrm commented on 2021-04-20 18:32 (UTC) (edited on 2021-04-20 23:11 (UTC) by marcosfrm)

This thing is dead even on Debian:

diff --git a/PKGBUILD b/PKGBUILD
index 1e8cfe7..8ea38d6 100644
@@ -55,7 +55,8 @@ package() {
        -e "s/x-scheme-handler\/ftp;\?//g" \

-   msg2 "Removing Debian Cron job and duplicate product logos..."
+   msg2 "Removing Debian Cron job, duplicate product logos and menu directory..."
    rm -r "$pkgdir"/etc/cron.daily/ "$pkgdir"/opt/google/chrome/cron/
-   rm "$pkgdir"/opt/google/chrome/product_logo_*.png
+   rm "$pkgdir"/opt/google/chrome/product_logo_*.{png,xpm}
+   rm -r "$pkgdir"/usr/share/menu/

angelettif commented on 2021-04-15 23:27 (UTC)

@yochananmarqos thank you. It was yup problem, I switched to paru

yochananmarqos commented on 2021-04-15 21:35 (UTC)

@angeletiff: Sounds like you have a faulty AUR helper.

angelettif commented on 2021-04-15 21:12 (UTC)

@yochananmarqos I cant, its downloaded automatically. DOnt know why it says: libunity (optional) , in my case its downloading it without question.

==> Installing Dependencies ==> Installing libunity 7.1.4-9 from the AUR ==> Making package: libunity 7.1.4-9 (jue 15 abr 2021 18:11:42) ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Retrieving sources... -> Found libunity_7.1.4+19.04.20190319.orig.tar.gz

angelettif commented on 2021-04-15 21:12 (UTC)

@yochananmarqos I cant, its downloaded automatically. DOnt know why it says: libunity (optional) , in my case its downloading it without question.

==> Installing Dependencies ==> Installing libunity 7.1.4-9 from the AUR ==> Making package: libunity 7.1.4-9 (jue 15 abr 2021 18:11:42) ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Retrieving sources... -> Found libunity_7.1.4+19.04.20190319.orig.tar.gz

yochananmarqos commented on 2021-04-15 16:45 (UTC)

@angelettif: Remove libunity, it's deprecated and unmaintained.

angelettif commented on 2021-04-15 15:52 (UTC) (edited on 2021-04-15 15:54 (UTC) by angelettif)

Cant upgrade: google-chrome 89.0.4389.114-1 -> 90.0.4430.72-2 Error on libunity compilation

protocol-icon.vala:188.3-188.27: error: overriding method `Unity.Protocol.AnnotatedIcon.serialize' is incompatible with base method `GLib.Variant? GLib.Icon.serialize ()': Base method expected return type `GLib.Variant?', but `GLib.Variant' was provided.
  private Variant serialize ()

yochananmarqos commented on 2021-04-13 22:24 (UTC)

As of v89, pipewire is used instead of libpipewire02.

luzifer commented on 2021-04-06 23:13 (UTC)

@ablecounter: fakeroot is part of base-devel.

From the Wiki:

Note: The group base-devel is assumed to be already installed when building with makepkg. Members of this group should not be included in makedepends array.

ablecounter commented on 2021-04-06 23:10 (UTC)

This seems to need fakeroot, so it sort of seems like that should be a dependency, unless I'm missing something obvious, which happens.

akhil commented on 2021-03-30 04:34 (UTC)

@revengel this worked for me $ pkill picom

akhil commented on 2021-03-30 04:26 (UTC)

@revengel same issue here. I posted it in forums also but no response.

Det commented on 2021-03-28 11:24 (UTC) (edited on 2021-03-30 16:00 (UTC) by Det)

@GeneralPoxter does that happen with extra/chromium (Built for Arch, not Debian/Ubuntu)?

E: Oh nvm, you said it's not a Chromium issue. Chrome (.deb, .rpm, whatever) is built for that one distro-set, so there's always gonna be some lib-mismatch issues due to their own versionings and distro-specific patches.

E2: Like even tho I loved maintaining this, I wouldn't bother with it for daily use.

GeneralPoxter commented on 2021-03-27 02:19 (UTC) (edited on 2021-03-27 04:53 (UTC) by GeneralPoxter)

Every now and then, google chrome would appear to timeout on reboot/shutdown:

Mar 26 22:07:35 poxter systemd[401]: app-gnome-google\x2dchrome-1873.scope: Stopping timed out. Killing.
Mar 26 22:07:35 poxter systemd[401]: app-gnome-google\x2dchrome-1873.scope: Killing process 1873 (chrome) with signal SIGKILL.
Mar 26 22:07:35 poxter systemd[401]: app-gnome-google\x2dchrome-1873.scope: Killing process 2192 (ThreadPoolSingl) with signal SIGKILL.
Mar 26 22:07:35 poxter systemd[401]: app-gnome-google\x2dchrome-1873.scope: Killing process 1903 (CrShutdownDetec) with signal SIGKILL.
Mar 26 22:07:35 poxter systemd[401]: app-gnome-google\x2dchrome-1873.scope: Killing process 1918 (ThreadPoolForeg) with signal SIGKILL.
Mar 26 22:07:35 poxter systemd[401]: app-gnome-google\x2dchrome-1873.scope: Failed with result 'timeout'.

Edit: I was able to reproduce this multiple times by simply rebooting, starting google chrome, and then rebooting again.

Edit 2: It appears to be good practice for the user to manually have to kill a browser before rebooting: . I've decided to turn off the "Continue running background apps when Google Chrome is closed" setting and will update if the problem persists.

Edit 3: It seems that even with background processes turned off, systemd timeouts still occur occasionally. I never encountered this issue in my year-long usage of chromium before I switched to chrome, so I'm not sure what's going on.

j1simon commented on 2021-03-19 14:58 (UTC)

@FabioLolix if you want to say something about this package do not use the "mark as obsolete" option. Use the comments, that's what they are there for. On top of that, your comment is not very relevant: "[i]gtk3-print-backends no longer exist and can be removed from optdepends[/i]".

marcosfrm commented on 2021-02-14 19:47 (UTC)

/usr/share/menu is as far as I know also obsolete on Debian. Please drop it.

nowy commented on 2021-02-05 02:31 (UTC)

I think /usr/share/menu folder should be deleted. It is necessary for debian menu system, not arch.

revengel commented on 2021-02-04 01:57 (UTC)

When watching video on 2x speed in some moment video picture accelerates to cosmic speed. Video sound speed is correct. After that video completely freezes, but webpage reload help.

The bug is reproduced on any google chrome version 88+.

After downgrading browser to version 87.0.4280.141 - browser works well.

The bug is reproduced on: * Google chrome v88.0.4324.96 * Google chrome v88.0.4324.146 * Google chrome unstable 90.0.4400.8 * Vivaldi 3.6.2165.34 (Chrome/88.0.4324.99)

Who did also see this bug? How I can fix it?

buzzra commented on 2021-01-31 02:02 (UTC)

ALL previous session/open tabs/open windows information was clobbered with this last update, Version 88.0.4324.96. Any possibility to recover?

akhil commented on 2021-01-23 14:25 (UTC)

smooth scrolling is broken in the latest update google-chrome-stable v88.0.4324.96

reverted to google-chrome-stable v87.0.4280.141 and smooth scrolling is working fine as it is.

IamJuras commented on 2021-01-19 16:11 (UTC)

In Gnome after upgrade to 87.0.4280.141-1 Chrome opens with default icon on dock and named "Unknown". In top it shows normal, as "chrome".

maxpayne3 commented on 2021-01-01 10:53 (UTC) (edited on 2021-01-01 11:01 (UTC) by maxpayne3)

Can you please add

sed -i "s/x-scheme-handler\/ftp;\?//g" "$pkgdir"/usr/share/applications/google-chrome.desktop

to fix this issue on Gnome?

You already change the desktop entry file with the following line in the PKGBUILD

sed -i "/Exec=/i\StartupWMClass=Google-chrome" "$pkgdir"/usr/share/applications/google-chrome.desktop

to fix the icon resolution on the dock in Gnome. Just add the other line, it's not so difficult. Thanks.

HunabKu commented on 2020-12-03 05:28 (UTC) (edited on 2020-12-03 05:54 (UTC) by HunabKu)

Hi, I had these error 4 times :

tput: No value for $TERM and no -T specified

So i set my terminal (alacritty) on /etc/environment and reboot. Now i have these error 4 times :

tput: unknown terminal "/usr/bin/alacritty"

The solution is to install xterm, set it as TERM in /etc/environment and reboot before reinstall google-chrome

lviggiani commented on 2020-11-11 10:26 (UTC)

WebRTC screen sharing seems now to be working again after receiving the update of pipewire 0.3.15-2

lviggiani commented on 2020-11-11 08:26 (UTC) (edited on 2020-11-11 08:27 (UTC) by lviggiani)

I've been using WebRTC screen sharing in Chrome + GNOME Shell + Wayland for long time (e.g. in Google meet calls) as described here:

Yesterday I recevide an update for Chrome (this package) and after that, screen sharing does not work anymore. It's not able to capture the entire screen. Does anyone know any workaroud? Thanks!

maxpayne3 commented on 2020-10-27 16:53 (UTC) (edited on 2020-10-27 19:00 (UTC) by maxpayne3)

According to this, can you please add the following command to exclude x-scheme-handler/ftp from the desktop entry? Thanks.

sed -i "s/x-scheme-handler\/ftp;\?//g" google-chrome.desktop

luzifer commented on 2020-10-25 16:14 (UTC)

@aryklein Providing the proper fonts for a website is a task for the websites creator, not for the system. Also I strongly doubt any Bootstrap page is using Liberation font:

The default font-family is "Helvetica Neue", Helvetica, Arial, sans-serif.

You're just providing a sans-serif font. So installing any sans-serif font will make things more beautiful than not providing any and relying on any font to be available for any reason.

aryklein commented on 2020-10-25 15:29 (UTC) (edited on 2020-10-25 15:30 (UTC) by aryklein)

Sites based on bootstrap4 look ugly if you don't install ttf-liberation which is an optional package.

luzifer commented on 2020-08-26 12:38 (UTC)

@j1simon which terminal do you use? Haven't seen any terminal not properly setting the TERM variable until now…

j1simon commented on 2020-08-26 07:00 (UTC)

Error in google-chrome.install:

(1/2) upgrading google-chrome                                                      [------------------------------------------------] 100%
tput: No value for $TERM and no -T specified
tput: No value for $TERM and no -T specified
tput: No value for $TERM and no -T specified
tput: No value for $TERM and no -T specified
==> NOTE: Custom flags should be put directly in: ~/.config/chrome-flags.conf
==> NOTE: The launcher is called: 'google-chrome-stable'

Ivellios commented on 2020-05-29 02:35 (UTC)

Not sure if this related to chrome or other package. Having issue using kwallet5 so switch to basic:

google-chrome-stable --password-store=basic

When saving password for some website: [26685:26957:0529/] Passwords sqlite error 2067, errno 0: UNIQUE constraint failed: logins.origin_url, logins.username_element, logins.username_value, logins.password_element, logins.signon_realm, sql: INSERT INTO logins (origin_url, action_url, username_element, username_value, password_element, password_value, submit_element, signon_realm, preferred, date_created, blacklisted_by_user, scheme, password_type, times_used, form_data, date_synced, display_name, icon_url, federation_url, skip_zero_click, generation_upload_status, possible_username_pairs, id, date_last_used) VALUES (?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)

ajkerzner commented on 2020-05-25 01:15 (UTC)

Please consider using HTTPS instead of HTTP for the package download.

ajkerzner commented on 2020-05-25 01:14 (UTC)

@Hendrikto Use $(< file) instead of $(cat <file). It's a minor simple thing.

xanadu73 commented on 2020-05-22 18:18 (UTC)

Works without issue on manjaro 20.1 thanx

Hendrikto commented on 2020-05-20 11:43 (UTC)

May I suggest applying this diff, to make the user flags XDG-compliant (see

---    2020-05-20 13:36:16.632332473 +0200
+++    2020-05-20 13:41:53.321282796 +0200
@@ -1,8 +1,10 @@

 # Allow users to override command-line options
-if [[ -f ~/.config/chrome-flags.conf ]]; then
-   CHROME_USER_FLAGS="$(cat ~/.config/chrome-flags.conf)"
+if [[ -f $XDG_CONFIG_HOME/chrome-flags.conf ]]; then
+   CHROME_USER_FLAGS="$(cat $XDG_CONFIG_HOME/chrome-flags.conf)"

 # Launch

j1simon commented on 2020-05-20 07:17 (UTC)

libunity is a library for a particular desktop environment that, in fact, has nothing to do with either KDE or Qt. Why then is it an optional dependency for download progress on KDE?

Det commented on 2020-05-11 11:46 (UTC) (edited on 2020-05-11 11:47 (UTC) by Det)

@tinywrkb, the only reason for me was the quicker out-of-date flags back when I used to be active and update often within minutes of emails (you could also --skipinteg, instead of users needlessly downloading old versions).

luzifer commented on 2020-05-10 19:41 (UTC)

@tinywrkb: Had no idea such a file exists… Just giving it a try - that would make the build more stable and better to maintain… Thanks for the hint!

tinywrkb commented on 2020-05-10 19:30 (UTC) (edited on 2020-05-10 19:31 (UTC) by tinywrkb)

Instead of having makepkg fail when the PKGBUILD is outdated, why not use a link to a specific version like this:
The only reason not to do this is if there are deb repackaging releases that could be missed, e.g. chrome-stable_81.0.4044.138-2_amd64.deb, chrome-stable_81.0.4044.138-3_amd64.deb.

yochananmarqos commented on 2020-04-27 23:19 (UTC)

@luzifer: 81.0.4044.129 is out:

@nikola135: Please do not flag packages out of date for build issues. Post things like that in the comments.

jaudriga commented on 2020-03-31 10:34 (UTC)

Screen Sharing on Wayland does not work without libpipewire02 on my machine. This is probably missing as an optional dependency like in .

nlehmann commented on 2020-02-26 21:29 (UTC)

Thanks, @Det and @neo2011, that's pretty useful information.

Det commented on 2020-02-26 18:29 (UTC)

PKGEXT=".pkg.tar.lzo" & PKGEXT=".pkg.tar.lrz" good compromises.

neo2001 commented on 2020-02-26 18:26 (UTC) (edited on 2020-02-26 18:32 (UTC) by neo2001)

Pretty sure this is normal, yes. I guess there are many small files. Check out /etc/makepkg.conf. You may want to set PKGEXT to a less aggressive and faster compression format (PKGEXT='.pkg.tar.gz') or don't compress the packages at all (PKGEXT='.pkg.tar')

nlehmann commented on 2020-02-26 18:11 (UTC)

Hi, I use pacaur to install my aur pacakges and noticed that whenever I update chrome it stays way too long in "Compressing package", much more than any other package. Is that normal?

pinbo commented on 2020-02-15 16:06 (UTC)

I am using Manjaro Gnome (5.5.2-1-MANJARO #1 SMP PREEMPT Tue Feb 4 19:22:35 UTC 2020 x86_64 GNU/Linux). Since last update, when Google Chrome is open, other windows under focus will lose focus after a few seconds. For example, when I was writing in Gedit, I cannot type in after a few seconds because the window has lost focus, and I need to click on it again to continue typing. Closing Google Chrome will solve the problem. Thanks.

arnohovhannisyan commented on 2020-02-09 07:31 (UTC)

As of v80.0.3987.87-1, I'm having problems with the maximize and restore buttons. When the window is not fullscreen, the maximize button doesn't do anything and it has the same icon as the restore button. This did not happen with v79. I'm using Manjaro GNOME if it matters.

felipea.braga commented on 2020-02-06 19:51 (UTC)

It seems to have a problem when using the "classic" theme on Manjaro KDE, has someone experienced this?

Det commented on 2020-02-06 13:54 (UTC)

echoing and sedding the thing with echo "diff [...] (END)" | sed "s/ +/\n+/g" | sed "s/diff \-\-/\n\ndiff \-\-/g" comes down to some patches:

So it's about python2-txaio (a fairly recent package, uploaded 2020-01-25):

"Al poner fin no realiza ninguna acción." auto-detects as [Spanish] in Google Translate and translates to "When ending it does not perform any action.":

jamezrin commented on 2020-02-06 13:31 (UTC)

@superhdj your comment is not very helpful, what happens? Also, speak english.

superhdj commented on 2020-02-05 23:33 (UTC)

diff --git /home/user/.cache/yay/python2-txaio/PKGBUILD /home/user/.cache/yay/python2-txaio/PKGBUILD new file mode 100644 index 0000000..a6a901a --- /dev/null +++ /home/user/.cache/yay/python2-txaio/PKGBUILD @@ -0,0 +1,34 @@ +# Maintainer: Jose Riha <jose1711 gmail com> + +pkgname=python2-txaio +pkgver=18.8.1 +pkgrel=1 +pkgdesc='Compatibility API between asyncio/Twisted/Trollius' +arch=('any') +url="" +license=('MIT') +makedepends=('python2-setuptools') +checkdepends=('python2-pytest' 'python2-mock' 'python2-twisted') +depends=('python2-six') +source=("$pkgver.tar.gz" + "make-pytest-happy.patch") + +prepare() { + # + patch -Np1 -d txaio-$pkgver <make-pytest-happy.patch + + cp -a txaio-$pkgver{,-py2} +} + +build() { + cd "$srcdir"/txaio-$pkgver-py2 + python2 build +} + +package() { + cd txaio-$pkgver-py2 + python2 install --root="$pkgdir" --optimize=1 + install -Dm644 LICENSE "$pkgdir"/usr/share/licenses/$pkgname/LICENSE +} +md5sums=('ecdd4d0f401c7f01f22830b4d2788593' + '862ac257bb05a8064b6e2814686dcb3c') diff --git /home/user/.cache/yay/python2-txaio/make-pytest-happy.patch /home/user/.cache/yay/python2-txaio/make-pytest-happy.patch new file mode 100644 index 0000000..dcf5e40 --- /dev/null +++ /home/user/.cache/yay/python2-txaio/make-pytest-happy.patch @@ -0,0 +1,47 @@ +From 9217f054b7eccc120f84e01995479125e07de59a Mon Sep 17 00:00:00 2001 +From: meejah <a href=""> +Date: Fri, 22 Feb 2019 12:01:46 -0700 +Subject: [PATCH] make pytest happy + +--- + test/ | 12 ++++++++++-- + 1 file changed, 10 insertions(+), 2 deletions(-) diff --git /home/user/.cache/yay/python2-txaio/PKGBUILD /home/user/.cache/yay/python2-txaio/PKGBUILD new file mode 100644 index 0000000..a6a901a --- /dev/null +++ /home/user/.cache/yay/python2-txaio/PKGBUILD @@ -0,0 +1,34 @@ +# Maintainer: Jose Riha <jose1711 gmail com> + +pkgname=python2-txaio +pkgver=18.8.1 +pkgrel=1 +pkgdesc='Compatibility API between asyncio/Twisted/Trollius' +arch=('any') +url="" +license=('MIT') +makedepends=('python2-setuptools') +checkdepends=('python2-pytest' 'python2-mock' 'python2-twisted') +depends=('python2-six') +source=("$pkgver.tar.gz" + "make-pytest-happy.patch") + +prepare() { + # + patch -Np1 -d txaio-$pkgver <make-pytest-happy.patch + + cp -a txaio-$pkgver{,-py2} +} + +build() { + cd "$srcdir"/txaio-$pkgver-py2 + python2 build +} + +package() { + cd txaio-$pkgver-py2 + python2 install --root="$pkgdir" --optimize=1 + install -Dm644 LICENSE "$pkgdir"/usr/share/licenses/$pkgname/LICENSE +} +md5sums=('ecdd4d0f401c7f01f22830b4d2788593' + '862ac257bb05a8064b6e2814686dcb3c') diff --git /home/user/.cache/yay/python2-txaio/make-pytest-happy.patch /home/user/.cache/yay/python2-txaio/make-pytest-happy.patch new file mode 100644 index 0000000..dcf5e40 --- /dev/null +++ /home/user/.cache/yay/python2-txaio/make-pytest-happy.patch @@ -0,0 +1,47 @@ +From 9217f054b7eccc120f84e01995479125e07de59a Mon Sep 17 00:00:00 2001 +From: meejah +Date: Fri, 22 Feb 2019 12:01:46 -0700 +Subject: [PATCH] make pytest happy + +--- + test/ | 12 ++++++++++-- + 1 file changed, 10 insertions(+), 2 deletions(-) +--- + test/ | 12 ++++++++++-- + 1 file changed, 10 insertions(+), 2 deletions(-) +++ /home/user/.cache/yay/python2-txaio/make-pytest-happy.patch @@ -0,0 +1,47 @@ +From 9217f054b7eccc120f84e01995479125e07de59a Mon Sep 17 00:00:00 2001 +From: meejah +Date: Fri, 22 Feb 2019 12:01:46 -0700 +Subject: [PATCH] make pytest happy + +--- + test/ | 12 ++++++++++-- + 1 file changed, 10 insertions(+), 2 deletions(-) + +diff --git a/test/ b/test/ +index 4a225c8..a3ca2d6 100644 +--- a/test/ ++++ b/test/ +@@ -25,9 +25,9 @@ def framework(request): + + try: + if request.param == 'twisted': +- return framework_tx() ++ return _notfixture_framework_tx() + elif request.param == 'asyncio': +- return framework_aio() ++ return _notfixture_framework_aio() + except ImportError: + pytest.skip() + +@@ -43,6 +43,10 @@ def framework_uninitialized(): + + @pytest.fixture + def framework_tx(): ++ return _notfixture_framework_tx() ++ ++ ++def _notfixture_framework_tx(): + try: + import txaio + from txaio import tx +@@ -56,6 +60,10 @@ def framework_tx(): + + @pytest.fixture + def framework_aio(): ++ return _notfixture_framework_aio() ++ ++ ++def _notfixture_framework_aio(): + try: + import txaio + from txaio import aio (END)

Al poner fin no realiza ninguna acción.

luzifer commented on 2020-02-03 00:35 (UTC) (edited on 2020-08-19 11:05 (UTC) by luzifer)

When reporting this package as outdated make sure there is indeed a new version for Linux Desktop. You can have a look at the "Stable updates" tag in Release blog for this. Do not report updates for ChromeOS, Android or other platforms stable versions as updates here.

This package will automatically get updated as soon as the Debian package is available. This is checked at least once per hour.

kidonng commented on 2020-01-15 06:49 (UTC) (edited on 2020-01-15 06:52 (UTC) by kidonng)

Should we add plasma-browser-integration as an optional dependency? Also, download progress works fine if the forementioned package is installed, so I assume libunity may be no longer needed.

jheuel commented on 2020-01-08 16:24 (UTC)

The segfaults from last version seem to be fixed. At least I got a full day without crashes.

jamezrin commented on 2019-12-19 23:27 (UTC)

Latest version segfaults randomly. I cleared its data folder and will see if that fixes it. Doesn't happen with chromium though. Just like @Serial I'm using deepin DE.

Serial commented on 2019-12-14 19:11 (UTC)

The new version 79.0.3945.79-1 of google-Chrome is closing the browser by itself.

I already disabled all extensions and still keeps closing alone.

Note: Chromium (chromium) 79.0.3945.79-1 is doing the same thing. Browser closes by itself, even though it has disabled all extensions.

The software version is Arcolinuxb-deppin version 19.11.3



Angaziwa commented on 2019-12-03 02:12 (UTC)

why the password can't be synced from other devices?and even click save websites' password,there still is empty?

Det commented on 2019-12-01 17:09 (UTC) (edited on 2019-12-01 17:10 (UTC) by Det)

@kirk, no they not required if they optional.

Not sure what u mean? KDE has KDE stuff, GNOME has GNOME stuff.

kirk commented on 2019-12-01 17:08 (UTC)

Can someone clarify why the following optional dependencies are are required, especially taken into account that not all users are using the same window managers or environments?

optdepends=('kdialog: for file dialogs in KDE' 'gnome-keyring: for storing passwords in GNOME keyring' 'kwallet: for storing passwords in KWallet' 'gtk3-print-backends: for printing' 'libunity: for download progress on KDE' 'ttf-liberation: fix fonts for some PDFs (CRBug #369991)' 'xdg-utils')

Lot's of kde stuff here and keyring software, are those even necessary at all?

luzifer commented on 2019-11-24 13:11 (UTC)

@rsevero the message you've posted looks to me like a broken package file… Just try to build the package again. (If you're using makepkg just delete the tar.xz file and rebuild)

rsevero commented on 2019-11-24 10:00 (UTC)

Having trouble updating to version 78.0.3904.108-1. After tons of warnings I get the following error:

(1/1) reinstalling google-chrome [########################################################] 100% error: could not extract /opt/google/chrome/chrome (Lzma library error: No progress is possible) error: problem occurred while upgrading google-chrome ==> NOTE: Custom flags should be put directly in: ~/.config/chrome-flags.conf ==> NOTE: The launcher is called: 'google-chrome-stable' error: could not commit transaction error: failed to commit transaction (transaction aborted) Errors occurred, no packages were upgraded.

And the worse is that now I have no Chrome at all as apparently the previous version was uninstalled.

Any ideas on how to get Chrome back?

chrisu commented on 2019-11-23 15:57 (UTC)

I have managed to resolve the issue with version 78 by deleting the contents of ~/.config/google-chrome and restarting. It appears to be a chrome bug and not specific to Arch.

luzifer commented on 2019-11-16 14:34 (UTC)

@quite hmm I'm not sure this should be done. In the end the packages requiring chromium are expecting to find a Chromium browser while this package contains a Chrome browser. It doesn't listen on the same binary (and to prevent collisions it shouldn't do so) and it behaves differently… So I think I'll refrain from adding this as basically Chrome is not Chromium…

quite commented on 2019-11-16 09:58 (UTC) (edited on 2019-11-16 09:58 (UTC) by quite)

@luzifer, would you consider provides=('chromium')? Or is that mad? It would help pkgs that work with any chromium-based browser.

C_Schmidpeter commented on 2019-11-14 11:09 (UTC)

I tried to update from v72 to v78, but I got the following error message:

error: could not extract /opt/google/chrome/nacl_helper (Lzma library error: No progress is possible) error: problem occurred while installing google-chrome

chrisu commented on 2019-11-12 22:27 (UTC)

The update to chrome 78 has also broken it for me. Running chrome starts a number of chrome processes, but no window appears and it hangs. Neither --disable-gpu, nor --disable-software-rasterizer worked.

As NeonCloud noted, sudo google-chrome-stable --no-sandbox allows Chrome to load and display, but it's far from ideal.

Unfortunately, I am unable to downgrade to version 77 as only 78 is available to download from Google. I have installed Chromium, which at the time of install was version 77, and I am able to use that for now.

Det commented on 2019-11-07 12:31 (UTC)

@NeonCLoud, did you try clean profile or Chromium that's actually built for Arch?

NeonCloud commented on 2019-11-03 03:44 (UTC) (edited on 2019-11-03 05:15 (UTC) by NeonCloud)

@birdmun: Can confirm, worked fine until a full system update recently and now it's completely broken. I've tried reverting mesa and libglvnd back several months to no avail. With --disable-gpu provided, it silently hangs. With --disable-software-rasterizer as well, it then provides a different error entirely.

Curiously: sudo google-chrome-stable --no-sandbox allows Chrome to load, but it's quite unstable, not connected to my user account, and seriously far from ideal!

Edit: Ok, I downgraded to 77.0.3865.120, and it's working again. The jump to v78 seems to have been the commit that killed it.

birdmun commented on 2019-10-24 09:43 (UTC)

Am I the only person having errors trying to run the latest stable version of Chrome? [714745:714745:1024/] Failed to create GL context with glXCreateNewContext. [714745:714745:1024/] gl::init::CreateGLContext failed [714745:714745:1024/] Could not create context for info collection. [714745:714745:1024/] gpu::CollectGraphicsInfo failed. [714745:714745:1024/] Exiting GPU process due to errors during initialization

Worse yet the process continues to run in the background without offering anything to interact with.

alanmartinpmp commented on 2019-09-29 23:18 (UTC)

Can I simply just say what a comfort to find somebody that really knows what they're discussing on the internet You definitely understand how to bring an issue to light and make it important. More and more people really need to read this and understand this side of your story. It's surprising you're not more popular since you certainly possess the gift.

luzifer commented on 2019-09-09 09:35 (UTC) (edited on 2019-09-09 09:35 (UTC) by luzifer)

@tomrlopes When installing I've just used the package as is without changing anything (built in a freshly created Docker container). The CastLabs Demos are not working for me either. What does is the Bitmovin Player Demo [1]. On that page you also should see widevine EME detected and the video should be playable.


tomrlopes commented on 2019-09-08 23:24 (UTC)

@ruario First I apologize I'm a noob. I have the file in /opt/google/chrome and in the aur git directory I use /home/tom/software/google-chrome/pkg/google-chrome/opt/google/chrome but the drm videos on won't play.

-rw-r--r-- 1 root root 7893124 Jun 2 19:07

luzifer commented on 2019-09-04 09:11 (UTC)

@det good question. I don't have any power to clean those comments here though I'd prefer to keep the comments clean of spam and it seems there is no anti-spam measure in place to keep the bots out… :(

Det commented on 2019-09-04 09:09 (UTC)

How did spam like take over AUR since I left and there's nothing to be like done about it?

ruario commented on 2019-09-02 04:39 (UTC)

@tomrlopes That is not a problem. It has never been listed in chrome://components/ on Linux because it does not use the components mechanism to update on Linux. It is just a bundled file. Look for within the package. If you want to see if it is working, go to and select one of the videos tagged with DRM.

tomrlopes commented on 2019-09-02 04:28 (UTC)

No widevinecdm in chrome://components/

kraftwerk28 commented on 2019-08-08 22:08 (UTC)

@yuu to enable dark mode just pass --force-dark-mode flag to the app

Det commented on 2019-07-16 10:44 (UTC) (edited on 2019-07-16 13:56 (UTC) by Det)

@TimOverboard, "trusted", because back when I maintained this thing the description used to be "An attempt at a faster, safer, more stable browser". Or something. Only so that the product would always stand on its own merit. Just, after a while people wanted change, so I made a poll and what-do-you-know this description won (see the very first pinned comment) :) (on top of which Chrome's got like 60% global usage share).

Either way, can't really see anything "inflammatory" about that, bro, nor does it somehow stigmatize luzifer as a troll to stick by a choice the majority decided on.

TimOverboard commented on 2019-07-16 10:26 (UTC) (edited on 2019-07-16 10:28 (UTC) by TimOverboard)

Description: The popular and trusted web browser by Google

The word trusted is unneccesary, subjective, promoting and inflammatory. Why would you state this as a fact? You are troll, yes? I sincerely hope so because the alternatives actually ruin my day.

I don't care how much longer you've been using arch and/or Linux than me.. just sort this out will you. It's grossly offensive to see it in my AUR search results.

JohnML commented on 2019-07-16 09:38 (UTC)

just updated my google-chrome to:

Google itself tells me: Version 75.0.3770.100 (Offizieller Build) (64-Bit)

but yay tells me it installed ok: google-chrome-75.0.3770.142-1

i used yay -Su

Google works ok, but are these two version different??

info: CPU: Quad Core Intel Core i5-6400 (-MCP-) speed/min/max: 800/800/3300 MHz Kernel: 5.1.16-1-MANJARO x86_64 Up: 26m Mem: 4018.2/15986.4 MiB (25.1%) Storage: 3.18 TiB (11.3% used) Procs: 234 Shell: bash 5.0.7 inxi: 3.0.34

Pamac refused to install: database locked

yuu commented on 2019-07-16 00:06 (UTC)

How to enable dark mode? I'm using Arch Linux and KDE plasma. GTK "Prefer Dark theme" is enabled but my google chrome still not in dark mode.

I've installed dark theme but I want dark mode. Address bar still white.

Help me. lol

jpvaldez26 commented on 2019-07-05 01:08 (UTC)

Flash not present on Chrome-Browser-Stable.

bianjp commented on 2019-05-17 02:02 (UTC)

The flash bug has been accepted by Chromium:

By the way, it's quite strange that Chromium still works with flash while Chrome does not.

imwxz commented on 2019-05-12 15:54 (UTC)

+1 flash not work. Temporary solution is to use

google-chrome-stable --no-sandbox

But you may get an ugly warning and may be not safe?

Oroggs commented on 2019-05-10 15:26 (UTC)

Flash cannot be used after the last version update

bparsons commented on 2019-04-24 12:33 (UTC)

The launcher script /usr/bin/google-chrome-stable in this package mangles flags that contain spaces or quotes. I fixed this but don't know where to submit the fix. Here is a gist with my adjustments:

luzifer commented on 2019-04-05 00:03 (UTC)

@kronikpillow like @Det already commented and pinned: In those cases just flag the package outdated...

kronikpillow commented on 2019-04-04 23:55 (UTC)

==> Validating source files with sha512sums... google-chrome-stable_73.0.3683.86_amd64.deb ... FAILED eula_text.html ... Passed ... Passed ==> ERROR: One or more files did not pass the validity check!

Det commented on 2019-04-02 01:07 (UTC)

@wingsuit, it's a binary pacakge, and hence mostly why. It was very much voted "trusted", and has +50% usage share in the world, but sure - not in the official repo. Just in Arch it's/used to be crashy, so Chromium, built for Arch, is better.

@jfcandidofilho, it was something like googling the "chrome deb/rpm repo" and prob also the URL in the /etc/cron.daily/ job that gets cleaned up from the package thru the line at the end of the PKGBUILD, which @luz also kept. I quit Arch & Linux, so my memory's a bit fading on these.

phief2 commented on 2019-03-30 15:30 (UTC)

Did ctrl+wheel zoom on websites stop working for anybody else along the way? It only works in PDF.js and other resource views now. Since I haven't used Chrome in a long time, I can't determine when the problem was introduced.

Perhaps it's just my configuration? kwin + x11 + synaptics driver.

Pryka commented on 2019-03-18 14:27 (UTC)

Is anyone have a strange download issue with v73.0.3683.75?

When I try to download a file bigger than ~100mb It just stops at 100% and nothing is happening, no matter how long I wait.

I tried in incognito, I restarted all Chrome settings, removed all extensions - nothing.

The only solution for me is to downgrade to v72.


wingsuit commented on 2019-03-13 09:51 (UTC)

"The popular web browser by Google (Stable Channel)"

I think we can all agree this software isn't trusted. It would be in the official repositories if it was.

jfcandidofilho commented on 2019-03-01 22:20 (UTC) (edited on 2019-03-01 22:25 (UTC) by jfcandidofilho)

@Det it's a bit unrelated to everything but where did you find those links for checksum and releases? I was googling about the checksum not working and if it weren't for you, I would fail to find a file with the checksum.

Of course, I could always direct myself towards the AUR but I sincerely couldn't find this info. I might just be blind...

Edit: I'm not well versed in packages functionality but we need an SHA512 and the file provides only SHA256, SHA1 and MD5sum. Is there a way to use MD5sum for that deb file only?

joelk commented on 2019-02-21 16:31 (UTC)

So, I uninstalled gnome-keyring and google-chrome-beta continued to work correctly. It loads and exits normally. Next I installed google-chrome-stable, tested to see that it worked correctly, then turned on Google Sync, and it crashed again just as it did before. I could get it to run by deleting .config/google-chrome but as soon as I turned on Sync again it crashed. Finally, I uninstalled libgnome-keyring (to do this I had to uninstall gksu, which I suppose I should have done anyway), started google-chrome-stable, turned on Sync and now it works correctly.

So to summarize, google-chrome seems to require that NEITHER gnome-keyring NOR libgnome-keyring be installed, or alternatively if libgnome-keyring is installed then gnome-keyring is needed temporarily when turning on Sync and it can subsequently be uninstalled. I have no explanation, but there it is.

leopicasso commented on 2019-02-21 09:26 (UTC)

Thanks for the information on the site. > Five Nights at Freddy's

luzifer commented on 2019-02-21 09:26 (UTC)

No XFCE, nothing containing gnome in the name. Using i3 as a window manager, avoiding packages belonging to Gnome or KDE.

joelk commented on 2019-02-21 03:30 (UTC)

Do you use xfce on any of them? Do you have libgnome-keyring (needed by gksu) on any of them?

luzifer commented on 2019-02-20 22:45 (UTC)

I do have neither gnome-keyring nor kwallet and all three variants of google-chrome do run quite fine on multiple machines of mine having turned on Google Sync on all of them. So there is no default dependency to those two packages.

joelk commented on 2019-02-20 22:04 (UTC)

I guess I can boil it down to a few key lines. I apologize if I'm spamming the board. I was able to run and close the program multiple times without starting google sync. But after I logged into a google account and started sync and then closed the browser, I could not re-open it. Each time that it ran successfully, the first 3 lines of terminal output were like these: mkdir: cannot create directory ‘/home/testuser/.local/share/applications’: File exists touch: cannot touch '/home/testuser/.local/share/applications/mimeapps.list': Not a directory Gkr-Message: 13:21:39.239: secret service operation failed: The name org.freedesktop.secrets was not provided by any .service files [17060:17060:0219/] InitializeSandbox() called with multiple threads in process gpu-process.

The last time (when it crashed) the org.freedesktop.secrets message was missing. The ONLY output was: mkdir: cannot create directory ‘/home/testuser/.local/share/applications’: File exists touch: cannot touch '/home/testuser/.local/share/applications/mimeapps.list': Not a directory [18518:18518:0219/] InitializeSandbox() called with multiple threads in process gpu-process.

Maybe the option is that it requires EITHER gnome-keyring or kwallet.

joelk commented on 2019-02-20 18:25 (UTC) (edited on 2019-02-20 18:28 (UTC) by joelk)

The crashes occurred yesterday, before I installed gnome-keyring. Now it's running fine -- I can load it and close it repeatedly without crashing. When it crashed, I didn't find any crash reports or any logs that seemed useful and afterwards I deleted the entire google-chrome directory. Then I installed google-chrome-beta and I did save a set of terminal outputs documenting several runs from initial installation until it crashed. The outputs appeared to be exactly the same as with the stable version. But it seems like a lot to post in these comments. I'd post it in the ArchLinux Forum, but when I tried posting just a small part of that yesterday the admin threw it in the dustbin because I wasn't "asking for support". Where do you suggest posting it?

Pryka commented on 2019-02-20 08:22 (UTC) (edited on 2019-02-20 08:27 (UTC) by Pryka)

@joelk Chrome should work without any kind of keyring installed on the system. So it is perfectly fine to have kwallet or keyring as an opt deps.

Run it via terminal and post some logs of this crash.

joelk commented on 2019-02-19 23:09 (UTC)

I think gnome-keyring should be a required, not optional, dependency. Without it, google-chrome crashes during start-up after sync has been turned on.

Det commented on 2019-01-06 11:26 (UTC)

@Alad, poll's over, bro.

Alad commented on 2019-01-06 00:32 (UTC)

I came for the spammers and stayed for the pointless polls.

diogobaeder commented on 2018-12-26 07:01 (UTC)

I'm getting an error with version 71.0.3578.98-1 when trying to enable fullscreen videos (e.g. YouTube or Vimeo): "TypeError: fullscreen error" - it seems to be related to this other error: "Failed to execute 'requestFullscreen' on 'Element': API can only be initiated by a user gesture."

It doesn't seem to be related to my DE, though, since fullscreen works fine in both Firefox and Chromium (version 71.0.3578.98-2).

Is anybody else affected by this?

Det commented on 2018-12-22 13:48 (UTC) (edited on 2018-12-22 13:48 (UTC) by Det)

@voot, it's built for Arch, not Ubuntu, which helps.

E: I can say that now because I don't maintain it. ( ͡° ͜ʖ ͡°)

voot commented on 2018-12-22 12:59 (UTC)

I used this for a year and half, though had an issue where no window would appear on startup though the chrome process was clearly present.

I'm going to try chromium which does not have this issue whatsoever.

(btw I start xfce with exec startxfce4 on every boot..., not sure if that was part of the issue, but whatever the case chromium runs fine)

Chromium really isn't that different than this after all...

Det commented on 2018-12-19 02:17 (UTC)

@bluedreamer, perhaps it's got a different compression or exported elsewhere as per makepkg.conf.

Should still find the package, but I don't even recall how that works.

luzifer commented on 2018-12-18 17:09 (UTC)

@bluedreamer I just tried to reproduce your error and wasn't able to. makepkg created exactly that package your makepkg is complaining about and installing it was successful… - I've no clue what went wrong between build and install in your output…

bluedreamer commented on 2018-12-18 15:19 (UTC)

Any idea why makepkg is failing?

user@kobold:~/code/google-chrome]$ makepkg -i ==> Making package: google-chrome 71.0.3578.98-1 (Tue 18 Dec 2018 08:17:36 AM MST) ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Retrieving sources... -> Found google-chrome-stable_71.0.3578.98_amd64.deb -> Found eula_text.html -> Found ==> Validating source files with sha512sums... google-chrome-stable_71.0.3578.98_amd64.deb ... Passed eula_text.html ... Passed ... Passed ==> Extracting sources... -> Extracting google-chrome-stable_71.0.3578.98_amd64.deb with bsdtar ==> Removing existing $pkgdir/ directory... ==> Entering fakeroot environment... ==> Finished making: google-chrome 71.0.3578.98-1 (Tue 18 Dec 2018 08:17:36 AM MST) ==> Installing package google-chrome with pacman -U... loading packages... error: '/home/user/code/google-chrome/google-chrome-71.0.3578.98-1-x86_64.pkg.tar.xz': could not find or read package ==> WARNING: Failed to install built package(s).

Det commented on 2018-11-24 12:37 (UTC)

Good-bye, friends.

I've left Arch a long time ago, and now I'm leaving these too.

To you.

ptkdev commented on 2018-11-18 00:21 (UTC)

Don't work Widevine plugin in latest version, example: netflix video crash.

yinnx commented on 2018-11-12 17:18 (UTC)

I have error: [2399:2399:1112/] Not implemented reached in virtual ui::InputMethodKeyboardController *ui::InputMethodBase::GetInputMethodKeyboardController()Using InputMethodKeyboardControllerStub

Error shows up when i interact with chrome ui and it lags for a second (opening menu or typing to address bar).

dschaefer79 commented on 2018-11-12 08:19 (UTC)

The Problem is gone again. lol

dschaefer79 commented on 2018-11-07 17:38 (UTC)

@spsf64 The problem appears again.

dschaefer79 commented on 2018-11-07 16:13 (UTC)

@spsf64 Installed with yaourt -S ice-ssb, it has solved my issue. Thanks.

Det commented on 2018-11-07 12:06 (UTC) (edited on 2018-11-07 12:11 (UTC) by Det)

Okay, guess they changed the format. I don't use Arch anymore, just update my remaining packages through Cygwin.

I can still check things like this also. I don't know at which point I drop my remainders, but just ping me about and we'll talk about it.

navarroaxel commented on 2018-11-07 11:44 (UTC) (edited on 2018-11-07 11:45 (UTC) by navarroaxel)

Looks like the AWK expression on

$ curl -sL | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }'

is not working anymore. I don't know why.

spsf64 commented on 2018-11-03 11:34 (UTC)

@dschaefer79 try to install ice-ssb: This will take care of necessary simlink for you. Hope it works

dschaefer79 commented on 2018-11-03 10:19 (UTC)

@tjuszki I have made what you said, but it's not working.

amansubhan commented on 2018-11-02 12:36 (UTC)

Installed the latest version last night, facing severe jerks and non fluid kinda software rendering.

sparky_005 commented on 2018-11-01 01:51 (UTC) (edited on 2018-11-01 02:06 (UTC) by sparky_005)

Anybody having issues with keyboard shortcuts not working on the latest version? Not sure if it's Chrome or something else.

EDIT: I fixed this by opening a new window from the menu and closing the existing windows. The existing windows, for some reason, weren't working, but new windows were. Weird bug!

tjuszki commented on 2018-10-26 15:40 (UTC)

@dschaefer79 In my case moving Google Chrome on top of the list in KDE Plasma -> System Settings -> Applications -> File Associations -> application/xhtml+xml resolved the issue.

dschaefer79 commented on 2018-10-26 15:30 (UTC)

I have again the problem set as default as the startup of chrome that won't go away. Does anyone have a solution ? Thanks,

Det commented on 2018-10-26 13:11 (UTC) (edited on 2018-10-28 12:39 (UTC) by Det)

zerophase flagged google-chrome out-of-date on 2018-10-26 for the following reason:

Chrome notifies an update is out.

Let it. There isn't.

Either it's an extension scamming you, or you don't have 70.0.3538.77. You have complete instructions right there how to check the latest versions. You could've even downloaded/verified the MD5 yourself.

antx commented on 2018-10-24 23:04 (UTC)

google-chrome-stable_70.0.3538.67_amd64.deb has wrong checksum.

Should be 297579a04aadce6ea9502e536e44edd3 according to PKGBUILD.

But on my machines I get this instead:

md5sum google-chrome-stable_current_amd64.deb ⏎ 2c51740867a6a8b5e53e227c62822465

dschaefer79 commented on 2018-10-24 12:29 (UTC)

The problem is gone

Det commented on 2018-10-21 13:00 (UTC) (edited on 2018-10-21 13:00 (UTC) by Det)

Depends very much on your DE (desktop environment, KDE, GNOME, etc.).

dschaefer79 commented on 2018-10-21 09:41 (UTC)

Everytime I start google-chrome it ask me about set as default, it can't set as default, what can I do ? Thanks,

Det commented on 2018-09-21 19:08 (UTC)

spicewiesel flagged google-chrome out-of-date on 2018-09-21 for the following reason:

md5 mismatch

No there isn't. Stop fake flagging.

The MD5 will only be taken from

Det commented on 2018-09-18 14:14 (UTC)

@jimreynold2nd, at that time when you tried.

@OrdoFlammae, I don't think there is a separate mirror. That's never been an issue, and trying just now, I connect directly to CA, USA (Silicon Valley).

@frydac, when you say this took you some time, you're not serious are you?

frydac commented on 2018-09-18 08:39 (UTC) (edited on 2018-09-18 08:41 (UTC) by frydac)

It did take me a little bit of time figuring this out, so hope this helps someone. Editing the pkgbuild with following entries worked for me:

version (using the "curl ..." command in the note by Det):

md5sum (downloading the .deb package manually and using md5sum):

OrdoFlammae commented on 2018-09-17 22:16 (UTC)

About the package being out of date, have you checked to make sure your mirror is updated?

jimreynold2nd commented on 2018-09-17 18:56 (UTC)

It is out of date: checksum does not match anymore.

Det commented on 2018-09-16 16:02 (UTC)

test1265tester flagged google-chrome out-of-date on 2018-09-16 for the following reason:

Its out of date

It isn't.

Det commented on 2018-09-12 10:14 (UTC) (edited on 2018-09-12 20:47 (UTC) by Det)

I reviewed my pinned comment, since people just can't fathom what's going on.

sku commented on 2018-09-11 22:19 (UTC)

md5sum of new version „69.0.3497.92“ is „58bafb3149bb0e0d00efe61357a3bd8a“ (first line of array). To be sure this should be reviewed.

Det commented on 2018-09-05 14:47 (UTC)

You noobs.

uncletrunks commented on 2018-09-05 05:32 (UTC)

The package isn't updated yet, hence the validity check fails. Edit the PKGBUILD if you can't wait, all you've gotta do is change the version number and MD5 for the .deb file.

riemannrocker commented on 2018-09-05 04:45 (UTC)

Another case of "One or more files did not pass the validity check!"

st8plgun commented on 2018-09-05 03:04 (UTC)

I'm getting the same error... "one or more files did not pass the validity check"

patrickcurl commented on 2018-09-05 02:43 (UTC) (edited on 2018-09-05 02:44 (UTC) by patrickcurl)

When upgrading, I keep getting: one or more files did not pass the validity check. (yay -Syu)

Enelar commented on 2018-08-31 05:37 (UTC)

@growler actually yes, it took several month to clean old crap

fix already merged to beta version, few more days dude

voot commented on 2018-08-29 06:12 (UTC)

Running on Xfce, also having the 'failing to fully close' issue.

The process needs to be killed manually, and the .config/google-chrome deleted for the window to appear/launch normally again.

Det commented on 2018-08-19 10:13 (UTC) (edited on 2018-08-19 10:15 (UTC) by Det)

You mean... the one that says:

  • Note: Git clone URLs are not meant to be opened in a browser.
  • To clone the Git repository of google-chrome, run git clone <>.
  • Click here to return to the google-chrome details page.

Hentaro commented on 2018-08-19 09:11 (UTC)

The git clone link is dead.

Det commented on 2018-08-11 19:15 (UTC)

New profile.

Pryka commented on 2018-08-01 17:51 (UTC) (edited on 2018-08-01 18:02 (UTC) by Pryka)

Does anyone have a problem with "zoomed-in" youtube videos(not all) in latest Chrome?

Found some old topic about this with pictures of the issue -!topic/chrome/jNnXAY41zP0

So far I thing only YT is affected. Can't see this behaviour on Netflix or HBO

Det commented on 2018-07-25 13:50 (UTC)

Or 68.0.3440.75 & 'SKIP'/--skipinteg.

windy commented on 2018-07-25 13:27 (UTC) (edited on 2018-07-25 13:27 (UTC) by windy)

diff for Chrome 68:



cedricmc commented on 2018-07-03 23:16 (UTC) (edited on 2018-07-03 23:17 (UTC) by cedricmc)

Issue #1: no system tray icon (Chrome, Hangouts). Issue #2: selecting "exit" in Chrome's taskbar dropdown menu does not fully close it and needs to be pkilled from a term.

Note: Gnome 3.28 is assumed.

Det commented on 2018-07-02 15:40 (UTC)

It's not made for Arch. ( ͡° ͜ʖ ͡°)

pbateman commented on 2018-07-02 15:40 (UTC)

Most popular browser and it's been broken for months now... Is it really that difficult to fix?

Det commented on 2018-06-27 11:46 (UTC)

The.. binaries here come straight from Google (built mostly for Ubuntu). The chromium package from [extra] is built for Arch using our libs.

bobarch64 commented on 2018-06-27 05:40 (UTC) (edited on 2018-06-27 05:41 (UTC) by bobarch64)

@Det, after reading more carefully: were chromium to adopt the current fontconfig library in its binary build that old config files would be read correctly here

how is the binary build for chrome? since chromium works fine with the new fontconfig library and conf

bobarch64 commented on 2018-06-21 07:26 (UTC)

yes it didn't work for me either. So meanwhile I will use chromium

Det commented on 2018-06-21 07:12 (UTC)

Well on the very next page someone else says it doesn't work even?

bobarch64 commented on 2018-06-21 07:10 (UTC) (edited on 2018-06-21 07:17 (UTC) by bobarch64)

@Det do you have any suggestions how to deal with it. Use the old fontconfig conf files?, as suggested in the and wait for a solution when chrome is updated (chromium is working fine btw)? anyway thanks for maintaining

Det commented on 2018-06-20 07:11 (UTC)

Not sure what do you expect to be done here, if LD_PRELOAD doesn't help even.

bobarch64 commented on 2018-06-20 05:44 (UTC)

Fontconfig outdated: it should be used the system fontconfig

ABOhiccups commented on 2018-06-18 08:45 (UTC)

Google Chrome won't open after installing Plymouth with arch-plymouthize when I turn on the computer.

Something has to do with /etc/fonts/fonts.conf Error.

topcat commented on 2018-06-12 18:28 (UTC)

md5sum failing as Google have updated to version 67.0.3396.87...

Det commented on 2018-06-02 10:19 (UTC) (edited on 2018-06-02 10:24 (UTC) by Det)

Well I've no idea what link u went into, but the blog (and direct link) are up to date.

You can always verify the version that gets pulled in directly with the repo data command, since the source link doesn't change.

E: The 66.0.3359.203 from May 31st, not 13th is for Chrome OS.

Det commented on 2018-05-29 19:47 (UTC)

Your package google-chrome [1] has been flagged out-of-date by newfell0w [2]:

release 66.0.3359.181 is current

Already updated.

Det commented on 2018-05-21 04:31 (UTC)

dnl_tp flagged google-chrome out-of-date on 2018-05-21 for the following reason:

google chrome asks to be updated... i guess it is outdated...

Nope.. see the pinned comment.

Det commented on 2018-05-15 19:01 (UTC) (edited on 2018-05-15 19:01 (UTC) by Det)

Your package google-chrome [1] has been flagged out-of-date by adherr [2]:

==> Validating source files with md5sums...

google-chrome-stable_66.0.3359.170_amd64.deb ... FAILED

It's already updated, mate.

Det commented on 2018-05-15 16:47 (UTC)

Nice spam from 2 weeks ago, didn't notice that.

Good thing they collapse now. <3

FXCNN commented on 2018-05-02 16:43 (UTC)

problemas al ejecutar google-chrome


google-chrome-stable Fontconfig warning: "/etc/fonts/fonts.conf", line 5: unknown element "its:rules" Fontconfig warning: "/etc/fonts/fonts.conf", line 6: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/fonts.conf", line 6: invalid attribute 'translate' Fontconfig error: "/etc/fonts/fonts.conf", line 6: invalid attribute 'selector' Fontconfig error: "/etc/fonts/fonts.conf", line 7: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/fonts.conf", line 7: invalid attribute 'version' Fontconfig warning: "/etc/fonts/fonts.conf", line 9: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/10-hinting-slight.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/10-hinting-slight.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/10-hinting-slight.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/10-hinting-slight.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/10-hinting-slight.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/10-hinting-slight.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/10-hinting-slight.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/20-unhint-small-vera.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/20-unhint-small-vera.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/20-unhint-small-vera.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/20-unhint-small-vera.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/20-unhint-small-vera.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/20-unhint-small-vera.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/20-unhint-small-vera.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/30-metric-aliases.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/30-metric-aliases.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/30-metric-aliases.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/30-metric-aliases.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/30-metric-aliases.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/30-metric-aliases.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/30-metric-aliases.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/40-nonlatin.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/40-nonlatin.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/40-nonlatin.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/40-nonlatin.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/40-nonlatin.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/40-nonlatin.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/40-nonlatin.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/45-generic.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/45-generic.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/45-generic.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/45-generic.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/45-generic.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/45-generic.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/45-generic.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/45-latin.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/45-latin.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/45-latin.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/45-latin.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/45-latin.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/45-latin.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/45-latin.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/49-sansserif.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/49-sansserif.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/49-sansserif.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/49-sansserif.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/49-sansserif.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/49-sansserif.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/49-sansserif.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/50-user.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/50-user.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/50-user.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/50-user.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/50-user.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/50-user.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/50-user.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/51-local.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/51-local.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/51-local.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/51-local.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/51-local.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/51-local.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/51-local.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/60-generic.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/60-generic.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/60-generic.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/60-generic.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/60-generic.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/60-generic.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/60-generic.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/60-latin.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/60-latin.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/60-latin.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/60-latin.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/60-latin.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/60-latin.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/60-latin.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/65-fonts-persian.conf", line 34: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/65-fonts-persian.conf", line 35: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/65-fonts-persian.conf", line 35: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/65-fonts-persian.conf", line 35: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/65-fonts-persian.conf", line 36: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/65-fonts-persian.conf", line 36: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/65-nonlatin.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/65-nonlatin.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/65-nonlatin.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/65-nonlatin.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/65-nonlatin.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/65-nonlatin.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/65-nonlatin.conf", line 8: unknown element "description" Fontconfig warning: "/etc/fonts/conf.d/69-unifont.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/69-unifont.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/69-unifont.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/69-unifont.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/69-unifont.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/69-unifont.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/80-delicious.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/80-delicious.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/80-delicious.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/80-delicious.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/80-delicious.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/80-delicious.conf", line 6: invalid attribute 'version' Fontconfig warning: "/etc/fonts/conf.d/90-synthetic.conf", line 4: unknown element "its:rules" Fontconfig warning: "/etc/fonts/conf.d/90-synthetic.conf", line 5: unknown element "its:translateRule" Fontconfig error: "/etc/fonts/conf.d/90-synthetic.conf", line 5: invalid attribute 'translate' Fontconfig error: "/etc/fonts/conf.d/90-synthetic.conf", line 5: invalid attribute 'selector' Fontconfig error: "/etc/fonts/conf.d/90-synthetic.conf", line 6: invalid attribute 'xmlns:its' Fontconfig error: "/etc/fonts/conf.d/90-synthetic.conf", line 6: invalid attribute 'version' Fontconfig error: Cannot load config file from /etc/fonts/fonts.conf

huiser commented on 2018-04-28 16:01 (UTC)

AUR Packages (1) google-chrome-66.0.3359.139-1

:: Proceed with installation? [Y/n] y :: Retrieving package(s)... :: Checking google-chrome integrity... ==> ERROR: PKGBUILD contains CRLF characters and cannot be sourced.

==> ERROR: An unknown error has occurred. Exiting... :: failed to verify google-chrome integrity

ohmysql commented on 2018-04-27 20:08 (UTC)

I get this message in build. I googled and couldn't see if it was bad :

==> Verifying database signature... ==> WARNING: No existing signature found, skipping verification. ==> Extracting database to a temporary location... ==> Verifying database signature... ==> WARNING: No existing signature found, skipping verification. ==> Extracting database to a temporary location... ==> Adding package 'google-chrome-66.0.3359.139-1-x86_64.pkg.tar.xz' -> Computing checksums... -> Removing existing entry 'google-chrome-66.0.3359.117-1'... -> Creating 'desc' db entry... -> Creating 'files' db entry...

foutrelis commented on 2018-04-19 17:32 (UTC)

@Pryka: Probably the same thing that happened to me: Chrome lost the random password it stores in the keyring as "Chrome Safe Storage", which it uses to encrypt cookies. After that is lost, any new cookies created don't overwrite their old counterparts and thus we end up with duplicate cookies.

The cookie handling behavior has changed in Chrome 66 and duplicate cookies should no longer be created (it was a rare issue in the first place). There might still be some problematic behavior though, as I noted in the upstream bug. It's also still the case that losing "Chrome Safe Storage" will make all cookies encrypted with it unreadable.

Pryka commented on 2018-04-19 16:58 (UTC)

@foutrelis Yeah, bunch of dupes.

But I don't understand now... Why creating a new profile fixed that on my machine? I jumped back and fort from 64 to 66 just to check it. And I'm not losing anything now. Chrome stays log in on sites an google account after each update.

foutrelis commented on 2018-04-19 16:48 (UTC)

@Pryka: If you have a backup of your Chrome 65 profile, you can run this query on its Cookies database:

If it returns anything other than 0, then duplicate cookies was the reason you lost your cookies after upgrading to Chrome 66.

Pryka commented on 2018-04-19 16:45 (UTC) (edited on 2018-04-19 16:45 (UTC) by Pryka)


Could be. But it seams to work normal after creating clean profile files. At least for now. Have to wait for the v67 to be sure.

foutrelis commented on 2018-04-19 16:37 (UTC)

@Pryke: Cookie loss after upgrading to M66 can be due to

@Det: You can remove gconf from the dependencies. It's not been needed since Chrome 65 I believe.

Pryka commented on 2018-04-19 16:26 (UTC)


Couldn't reproduce it in Chromium 65 and 66.

Apparently it was some kind of corruption of my Google-Chrome profile. Created new one. Downgraded/Updated(65<-->66) and everything is working fine now(with sync on and off). Thank You.

Pryka commented on 2018-04-18 19:51 (UTC) (edited on 2018-04-18 19:52 (UTC) by Pryka)

@Det Thank you for the interest.

I'll do this tomorrow, anyway I will get back to You and post results.

Det commented on 2018-04-18 19:37 (UTC)

It'd be interesting, if you could test this with Chromium also:

And if can't reproduce, then with clean profile on Chrome (non-synced) (see /var/cache/pacman/pkg or e.g.

Pryka commented on 2018-04-18 18:51 (UTC) (edited on 2018-04-18 19:01 (UTC) by Pryka)


Probably since v64 every time I perform chrome update to next major version like 64-->65 or now to 66. Chrome is losing all my sessions and I need to relog to almost every single site I use. Not ground breaking but annoying.

Anyone noticed similar behaviour?

PS. This bug or whatever it is do not occur when update are performed is in the same line, like: 65.0.3325.162 --> 65.0.3325.181.

EDIT: Did a little digging and it's looks that I have lost a whole lot of cookies after update. So this was probably the culprit in the previous ones(updates). Still no idea why this is happening.

Det commented on 2018-04-18 16:39 (UTC)

Oh my god, don't post a comment about it.

glowingsword commented on 2018-04-18 13:50 (UTC)

Please, update the PKGBUILD, the release to


and md5sums to

md5sums=('c979c8fac57b9f78de0b899f1b09ce39' 'd50d8f0a6940791eabc41c4f64e6a3cf' '99fa93d5e7fb5d622cef0f9621f3ffa3')

zerophase commented on 2018-04-17 00:09 (UTC)

@fuan_k Thanks. I just needed to confirm my overclocked ram wasn't causing the segfault.

fuan_k commented on 2018-04-16 23:39 (UTC)

@zerophase: yes, got the same segfault with 390.48 drivers, downgraded to 387.34.

zerophase commented on 2018-04-14 16:53 (UTC)

Has anyone else gotten this segfault?

[Sat Apr 14 11:41:35 2018] chrome[6358]: segfault at 4 ip 00007ff51a018967 sp 00007fff116c9388 error 6 in[7ff518df5000+141e000]

Det commented on 2018-04-02 14:17 (UTC)

Smoerrebroed flagged google-chrome out-of-date on 2018-04-02 for the following reason:

New release is out!

No there isn't.

v8areu commented on 2018-03-26 00:59 (UTC) (edited on 2018-03-26 01:00 (UTC) by v8areu)

So for now, we cannot log in and sync the extension etc for it to be run properly? Because (as others have mentioned below) if I ever log in + sync, google chrome won't start.

I don't actually need this to be synced because I mainly use chromium and I want to use chrome for testing purpose. But this is weird.

Det commented on 2018-02-25 12:01 (UTC)

Wow, yes it is.

That's super sad and annoying.

TiD91 commented on 2018-02-25 10:19 (UTC)

Ok. I just saw pacaur is unmantained... Arch is too fast for me sometimes :)

Det commented on 2018-02-25 10:18 (UTC)

clear ~/.cache/pacaur

TiD91 commented on 2018-02-25 10:17 (UTC)

Guys, If I try to install this with pacaur, I get: :: Installing google-chrome package(s)... :: google-chrome package(s) failed to install. :: ensure package version does not mismatch between .SRCINFO and PKGBUILD :: ensure package name has a VCS suffix if this is a devel package

But installing it manually works ok. Any idea how to fix this?

Det commented on 2018-02-23 18:13 (UTC)

Manjaro? You'll be doing another perfect update in a minute.

Jorre commented on 2018-02-23 16:22 (UTC)

I also ran into the md5 checksum error: ==> Validating source files with md5sums... google-chrome-stable_64.0.3282.167_amd64.deb ... FAILED

I decided to run yaourt with skipping validity checks: yaourt --m-arg "--skipchecksums --skippgpcheck" -Syua

This works fine for me and chrome updated perfectly.

Det commented on 2018-02-06 12:46 (UTC)


I just use pkill.

beoldhin commented on 2018-02-06 09:51 (UTC)

Added a Github page describing the shutdown problems observed with chrome and chromium (includes fix to "Chrome didn't shutdown correctly"):

rakotomandimby commented on 2018-02-05 19:05 (UTC)

You're right, the synchronization screws it all. If I dont sync, it restarts all right. If I ever sign in, it syncs then I cant relaunch it unless I remove config.

Workaround it to stay signed out for the moment. But that's weird.

I'll configure synchronization one by one and tell.

Det commented on 2018-02-05 18:59 (UTC)

Probably you're not closing it right. Or it happens after syncing extensions, etc.

I'd go step by step.

rakotomandimby commented on 2018-02-05 18:44 (UTC)

But it only works 1 launch after the remove of the directory: if I close Chrome and relaunch again, it fails.

rakotomandimby commented on 2018-02-05 18:41 (UTC)

"rm -rf ~/.config/google-chrome ; rm -rf ~/.cache/google-chrome" made it works for me...

Det commented on 2018-02-03 13:21 (UTC) (edited on 2018-02-03 13:22 (UTC) by Det)

What else is there besides the sums?

MarcinWieczorek commented on 2018-02-03 09:40 (UTC)

Please retire md5 and switch to sha256 at least for checksums :)

Det commented on 2018-02-02 03:13 (UTC) (edited on 2018-02-02 03:13 (UTC) by Det)

@belochub: for the differences between those, see:

The other two are:,

fuan_k commented on 2018-01-23 19:16 (UTC)

@clepi: from my experience, it's Chrome populating its History information or something. You can see it read a lot of things from disk and load it in memory. Can be databases generated by some extensions too, mostly related to history and cache usually.

ciepi commented on 2018-01-21 15:26 (UTC)

I have cleared ~/.config/google-chrome & ~/.cache/google-chrome but it still won't work.

Det commented on 2018-01-21 14:52 (UTC)

Went through the normal troubleshooting of renaming ~/.config/xxx/ & ~/.cache/xxx/, and looking at the terminal output?

ciepi commented on 2018-01-21 14:48 (UTC)

When I open chrome (or even chromium) I cannot use it - it's not responsive (can't even type anything in search bar). It've been installed without errors. Does anyone have the same issue? Moreover it works perfectly fine on my laptop, but it won't on pc.

beoldhin commented on 2018-01-17 08:51 (UTC) (edited on 2018-01-18 21:25 (UTC) by beoldhin)

I get the message "Chrome didn't shutdown correctly" always when I start Chrome after restarting the computer. I found out that Chrome leaves some processes running even after the browser is closed. Closing these processes with "killall chrome" before restart fixes this problem. The default signal sent by killall is "sigterm" but systemd probably sends something "faster" to the processes during shutdown, leaving them in unclean state.

Update: grepping chrome process tree root process with "pgrep -o chrome" and then killing that process + waiting for the directory to disappear from proc works better than "killall" that sometimes works, sometimes doesn't.

zerophase: "Is anyone else having issues with Chrome refusing to launch lately?". Yes. I just moved from chromium to chrome in one system with Cinnamon and other with Gnome 3. Launching chrome in Cinnamon doesn't work (hangs forever) but in Gnome 3 it does so there does seem to be that almost required dependency to gnome-keyring.

zerophase commented on 2018-01-12 06:32 (UTC) (edited on 2018-01-12 07:03 (UTC) by zerophase)

Is anyone else having issues with Chrome refusing to launch lately?

Just getting this error:

Gkr-Message: secret service operation failed: The name org.freedesktop.secrets was not provided by any .service files

EDIT: Well, I just had to install gnome-keyring. Have no idea what changed requiring that for Chrome to run.

xgdgsc commented on 2017-12-26 05:02 (UTC)

I also see severe memory usage issue in recent versions. I open hundreds of tabs and use tab suspender. It used to work with only less than 2GB, now I have to restart it once every day to free 6GB of memory or it will take all of swap.

angelv commented on 2017-12-20 21:26 (UTC) (edited on 2017-12-20 22:05 (UTC) by angelv)

I see. Didn't realize about the previous high memory comment. I guess it is probably the same thing. I don't know, since I was not monitoring memory usage and the computer became truly unusable even after waiting for a few minutes, so I couldn't even change to a terminal. I will check today and see if the memory is also the issue in my case.

After doing some tests, my issue seems different. Memory usage was a total of about 1.7GB (in a system with 8GB) with 8 tabs open. I tried to resize the Chrome window and suddenly the whole thing froze....

Det commented on 2017-12-20 21:09 (UTC) (edited on 2017-12-20 21:09 (UTC) by Det)

As in like as per the previous comments?

angelv commented on 2017-12-20 21:08 (UTC)

Anybody experiencing troubles with the latest google-chrome-stable? I updated the system during this week, and now I cannot use chrome properly: it works OK for some time, but at some point the computer freezes (I can move the mouse) but cannot click on anything and the keyboard is also unresponsive and I have to hard boot. I'm not completely sure chrome is responsible, but it has always happened (at least 5-6 times in a few hours) when I was using it and none when I have switched to Firefox.

Det commented on 2017-12-13 16:12 (UTC) (edited on 2017-12-13 16:12 (UTC) by Det)

Previous one:

Also see /var/cache/pacman/pkg/ (it's good to move AUR packages there also).

pazof commented on 2017-12-13 16:08 (UTC)

Also need to report the memory issue with 63.0.3239.84-1 - it now eats up at least 3 times the memory it used to - running it on a 16GB RAM system, the computer can hardly operate without lagging after opening the 5th Chrome tab or so.

DeadMetaler commented on 2017-12-09 09:53 (UTC)

google-chrome 63.0.3239.84-1 has memory leak for me. After some time it eats 4GB of memory, when normal for my situation is 1GB.

Det commented on 2017-11-25 01:29 (UTC) ?

znmeb commented on 2017-11-25 01:27 (UTC)

Chrome throws an error when starting from the command line: $ google-chrome-stable /opt/google/chrome/google-chrome: symbol lookup error: /usr/lib/ undefined symbol: FcWeightToOpenType Note that Chromium throws the same error, but I'd rather have Chrome working ;-)

cswl commented on 2017-11-20 13:06 (UTC)

I see, I use XFCE, but I might have pulled some gnome depedencies. I didn't notice the optdeps section during building. Sorry for the trouble.

Det commented on 2017-11-19 11:10 (UTC)

AFAIK it's a dependency only on GNOME. That's why it's an optdep.

cswl commented on 2017-11-19 09:06 (UTC)

Update.. I launched it from terminal and got this error.. Gkr-Message: secret service operation failed: The name org.freedesktop.secrets was not provided by any .service files Installing "gnome-keyring" it launched and asked to set up keyring.. So maybe add dependency on gnome-keyring?

cswl commented on 2017-11-19 08:48 (UTC)

Google chrome doesn't start up after I signed in into it.

Det commented on 2017-11-17 09:00 (UTC) (edited on 2017-11-17 09:00 (UTC) by Det)

> ccheek21 flagged google-chrome out-of-date on 2017-11-17 for the following reason: > > md5sum doesn't match It does.

bazzawill commented on 2017-11-14 03:00 (UTC)

Nevermind, pacaur didn't find a update but I just manually updated. User error sorry to bother

Det commented on 2017-11-11 09:05 (UTC) (edited on 2017-11-11 14:28 (UTC) by Det)

Gdb? Strace? Since when is this? Does a clean ~/.config/google-chrome & ~/.cache/google-chrome help?

Det commented on 2017-10-11 00:37 (UTC)

I'm assuming you mean /opt/google/chrome/google-chrome, but my line 49 says: exec -a "$0" "$HERE/chrome" "$@"

horosgrisa commented on 2017-10-10 20:33 (UTC)

file /usr/bin/google-chrome-stable have a bug. Please add a space between "$HERE/chrome" and "$CHROMIUM_USER_FLAGS" at line 49.

Det commented on 2017-10-07 12:59 (UTC)

I dunno. Looks like I don't need to add it, but can do anyway:,

ruhly commented on 2017-10-06 21:45 (UTC)

it would be useful to add gtk3-print-backends as an optional dependency for printing

Det commented on 2017-10-01 00:35 (UTC) (edited on 2017-10-01 00:42 (UTC) by Det)

The final vote. “What should be the pkgdesc of AUR/google-chrome?” With 82 votes: ● "One fast, simple, and secure browser for all your devices (Stable Channel)" - 17 votes (21%) ● "The popular and trusted web browser by Google (Stable Channel)" - 65 votes (79%) Now, my deciding vote of 49 votes will be cast and the winner is..., I’m kidding, I’m kidding. Pkgdesc changed. ^^

zman0900 commented on 2017-09-28 01:31 (UTC)

@ratan: `rm ~/.cache/pacaur/google-chrome/eula_text.html` should fix it.

ratan commented on 2017-09-27 19:30 (UTC)

getting the following when building from pacaur: :: google-chrome package(s) failed to install. :: ensure package version does not mismatch between .SRCINFO and PKGBUILD :: ensure package name has a VCS suffix if this is a devel package

Det commented on 2017-09-27 02:40 (UTC)

libc++'s not required by google-chrome?

fuan_k commented on 2017-09-26 23:57 (UTC)

If you get the following errors (while building libc++) ==> Verifying source file signatures with gpg... llvm-5.0.0.src.tar.xz ... FAILED (unknown public key 0FC3042E345AD05D) libcxx-5.0.0.src.tar.xz ... FAILED (unknown public key 0FC3042E345AD05D) libcxxabi-5.0.0.src.tar.xz ... FAILED (unknown public key 0FC3042E345AD05D) ==> ERROR: One or more PGP signatures could not be verified! It seems that a way to fix it is to run the following command (ref: gpg --recv-keys B6C8F98282B944E3B0D5C2530FC3042E345AD05D

Det commented on 2017-09-26 08:32 (UTC) (edited on 2017-09-26 08:42 (UTC) by Det)

Actually, about that, those who might find it interesting or entertaining:

Det commented on 2017-09-25 09:48 (UTC) (edited on 2017-09-26 15:12 (UTC) by Det)

That's a misconception between preimage resistance and collision resistance: Collision basically means finding/generating any 2 strings/files that have the same hash. For MD5 since 2013 that seems to be 2^18, which "runs in less than a second on a regular computer". Preimage is having that one predetermined file, for which you need to find/generate a string/file to match. Since 2009 that sems to be 2^123.4, which is... quite a lot more. E: What am I doing. I linked the same page you did. Sublinked.

JelteF commented on 2017-09-25 09:15 (UTC)

Could the checksums in this package please be changed to sha256 instead of md5? MD5 hasn't been a secure hash function since a long time now:

Det commented on 2017-09-25 05:26 (UTC)

> Your package google-chrome [1] has been flagged out-of-date by ytarch [2]: > > because md5sum is invalid. I can't build package. lol

Det commented on 2017-09-24 17:51 (UTC) (edited on 2017-09-24 17:51 (UTC) by Det)

> Your package google-chrome [1] has been flagged out-of-date by ytarch [2]: > > invalid md5sum. No it isn't.. Clear your cache.

Det commented on 2017-09-23 17:45 (UTC) (edited on 2017-09-23 17:45 (UTC) by Det)

> Your package google-chrome [1] has been flagged out-of-date by lookcrabs [2]: > > ensure package version does not mismatch between .SRCINFO and PKGBUILD Read damn comments.

Det commented on 2017-09-23 17:43 (UTC)

Probably due to the now bundled 'eula_text.html'. Very confusing message.

vstuart commented on 2017-09-23 17:40 (UTC)

@frimen: thank you! Same issue / solution! :-)

BurhanDanger commented on 2017-09-23 10:38 (UTC)

If you're facing problem like "frimen" noted , do a clean build with makepkg, you'll be fine. That was probably a pacaur issue.

grinderz commented on 2017-09-23 08:54 (UTC) (edited on 2017-09-23 11:53 (UTC) by grinderz)

:: google-chrome package(s) failed to install. :: ensure package version does not mismatch between .SRCINFO and PKGBUILD :: ensure package name has a VCS suffix if this is a devel package UPD: doing this 'rm -rf ~/.cache/pacaur/google-chrome' fix it. Thanks.

Det commented on 2017-09-22 21:03 (UTC) (edited on 2017-09-26 15:15 (UTC) by Det)

Heh. Here's a poll: (closes at 2017-09-29 21:03) 1) "The popular and trusted web browser by Google (Stable Channel)" 2) "One fast, simple, and secure browser for all your devices (Stable Channel)"

Det commented on 2017-09-22 17:49 (UTC)

I think I might be willing to go something like "The popular and trusted web browser by Google".

Det commented on 2017-09-22 17:43 (UTC)

An attempt at minimal design with sophisticated technology. lol

technic93 commented on 2017-09-22 15:42 (UTC)

Well, in deb they have Description: The web browser from Google Google Chrome is a browser that combines a minimal design with sophisticated technology to make the web faster, safer, and easier.

Det commented on 2017-09-22 15:29 (UTC)

Maybe. But. If using my own words and not even their official phrasing, it would sound more like it's tantamount to just an opinion. I've always liked how the thing stands on its own merit.

petershinners commented on 2017-09-22 15:25 (UTC)

Perhaps it is time to change the summary? I don't believe "attempt" is valid any longer. Without getting creative it could be shortened to: A safe, fast, and stable browser developed by Google (Stable Channel)

ydallilar commented on 2017-09-22 14:20 (UTC)

:: google-chrome package(s) failed to install. :: ensure package version does not mismatch between .SRCINFO and PKGBUILD :: ensure package name has a VCS suffix if this is a devel package

naibafo commented on 2017-09-19 13:05 (UTC)

For everybody struggling with the installation of the newest update that use a caching AUR helper like 'pacaur'. You need to delete the repository-folder from the cache because eula_text.html was is now part of the repository but also still exists in an untracked version in your local cache. For 'pacaur': 'rm -rf ~/.cache/pacaur/google-chrome'

Det commented on 2017-09-18 16:32 (UTC)

Seems there's some of those around: - - - If you reinstall and get the same thing, it's almost surely a false positive.

Tondo commented on 2017-09-18 15:59 (UTC)

ClamAV just threw /opt/google/chrome/chrome: Unix.Trojan.Mirai-5932143-0 FOUND

leondevlifelog commented on 2017-09-18 07:26 (UTC)

thanks,i has delete it(google-chrome-cn)

Det commented on 2017-09-17 17:56 (UTC) (edited on 2017-09-17 18:07 (UTC) by Det)

Uhhh. Yea, I rather would, because that one will be deleted: You need to have something notable change in the package itself to create a separate one. Different source is not enough. :( E: Added.

leondevlifelog commented on 2017-09-17 12:40 (UTC)

could you ues offline lisence file? I am in China,due to GFW,I can download deb file,but i can not download eula_text file,it show time out error.

Det commented on 2017-09-14 19:24 (UTC)

Shouldn't have anything to do with package signing. I just updated the package.

SACViper commented on 2017-09-14 19:23 (UTC)

I was getting was getting error ... Passed ==> ERROR: One or more files did not pass the validity check! Solution: I updated keys with this command and got installed without any error :) sudo pacman-key --refresh-keys sudo pacman-key --populate archlinux manjaro

darkxsun commented on 2017-09-07 00:03 (UTC)

@WorldwoWires probably relevant:

WorldwoWires commented on 2017-09-06 23:45 (UTC)

Is it just me or is something up with the color in google-chrome 61.0.3163.79-1? Seems to be washed out. Same with Chromium. It's definitely not in the previous versions after rolling back and updating and rolling back. What gives?

kirillbobyrev commented on 2017-09-05 22:12 (UTC) (edited on 2017-09-17 13:24 (UTC) by kirillbobyrev)

Not sure whether this issue is caused by the package or there is any other reason, but I wonder whether there is a way to resolve this issue. The system taskbar icon is looking great, but the tray icon is indeed resolution. Is tehre any way this could be fixed?

commented on 2017-09-01 07:13 (UTC)

constantly get "aw snap" crashes while watching "youtube tv". not recommended.

Det commented on 2017-08-26 03:58 (UTC)


whu_hyzs commented on 2017-08-26 03:55 (UTC) (edited on 2017-08-26 06:06 (UTC) by whu_hyzs)

Fcitx can't input Chinese on Chrome with GTK3(HiDPI) environment. Finally, I find the solution. First, moidfy /usr/bin/google-chrome-stable, add "export GDK_SCALE=1" before "exec /opt/google/chrome/google-chrome... ..." Second, touch a new file "~/.config/chrome-flags.conf" if not exist, add "--force-device-scale-factor=2" as parameters.

Det commented on 2017-08-20 16:56 (UTC)

Sort of. Only advantage is, I get the out-of-date flags quicker. I can add it to the PKGBUILD tho. So far it's only in -dev:

grawity commented on 2017-08-20 16:53 (UTC)

Instead of the "current" URL, wouldn't it be better to download the specific version from the apt-get repository? source=("${_channel}/google-chrome-${_channel}_${pkgver}-1_amd64.deb" (I had to use this when temporarily downgrading to v59.)

Det commented on 2017-08-14 18:23 (UTC)

Because this is the repository for super nerd Archers who build almost each of these packages themselves, causing so much makedependency overlap that we've had to create a whole entire group just for that. We call it the base-devel. It includes 25 of the most common build-time dependencies (24 + pacman), and with that in mind, we don't have to include any of those packages in our PKGBUILDs; we assume you already have the whole thing: - - - This is also discoverable by clicking the *All comments* and Ctrl+F'ing the "fakeroot".

joyarzun commented on 2017-08-14 15:31 (UTC)

-> Extracting google-chrome-stable_60.0.3112.90_amd64.deb with bsdtar ==> Removing existing $pkgdir/ directory... ==> Entering fakeroot environment... /usr/bin/makepkg: line 170: fakeroot: command not found ==> ERROR: Makepkg was unable to build google-chrome. ==> Restart building google-chrome ? [y/N] Why fakeroot is not a dependency??? I just installed manjaro gnome flavor

voot commented on 2017-07-19 02:33 (UTC) (edited on 2017-07-19 02:33 (UTC) by voot)

Noting here for others: I had an issue where manually installed microsoft fonts were causing 'aw snap' crashes. It seems like the crashes only started with 59 as I didn't have the same crashes before that. Also, the 'aw snap' crashes occurred independent of my using stable/beta/dev channels installed via aur.

ppc52776 commented on 2017-06-18 18:48 (UTC)

If you have CompositorTileW problem, try to remove font 'opendesktop-fonts'. [ 1402.814324] CompositorTileW[6608]: segfault at 39 ip 00005609d8ce6af0 sp 00007f2ec1f19068 error 6 in chrome[5609d81ba000+6a81000]

fuan_k commented on 2017-06-17 17:46 (UTC) (edited on 2017-06-17 17:47 (UTC) by fuan_k)

@TID91: isolate which versions are breaking, and report to the Chromium project? I doubt that has anything to do with the way Chrome is packaged for Arch (but I could be wrong).

TiD91 commented on 2017-06-16 20:36 (UTC)

Hello, I'm experiencing a problem with Chrome: while uploading a fair amount of photos on Google Photos (over 1000 photos) or videos (45 videos), after two or three files processed Google Chrome gets stuck and the popup "Page unrensposive" comes out and I need to kill chrome. I tried with Chromium on the official repositories and I have no problem with it. Any idea why? Do you have any suggestion on how to debug this?

valandil commented on 2017-06-10 18:27 (UTC)

@Det I do, but after resigning it I had lost my Session Manager info. I'll try with google-chrome-beta to see if it repopulates that info.

Det commented on 2017-06-10 08:00 (UTC)

@valandil, don't you sync your profiles with your Google account? commented on 2017-06-10 05:43 (UTC)

At the first start it is absolutely always chrome crashes, absolutely always. Such problems began with this version

valandil commented on 2017-06-10 01:03 (UTC)

@Det: Indeed, this works. I'll try to seek some help in the Chrome help forums, as I'd rather not lose my profile info. Thanks!

Det commented on 2017-06-09 19:07 (UTC)

Can't say. Someone in there also had an i7-3770. If Beta Channel works (60.x), then maybe you just need to stick to that branch, until Stable syncs up. Beta Channel also has its own settings dir, so maybe try a clean new profile by renaming ~/.config/google-chrome/ and ~/.cache/google-chrome/?

valandil commented on 2017-06-09 18:12 (UTC)

Have you ever seen this error? I've come accross this in old bugs (, but I think they all pointed to the use of old processors that didn't support SSE2. I'm on an Intel Core i7 which definitely supports SSE2, and still have this mistake. google-chrome-beta starts up just fine.

Det commented on 2017-06-06 14:21 (UTC)

Yeah the Beta and Dev's are too, just didn't notice the bump to 59: - -

lonaowna commented on 2017-06-06 07:47 (UTC)

sflor is right. see also:

sflor commented on 2017-06-06 07:36 (UTC)

Dependency change? gtk2 -> gtk3

Det commented on 2017-05-28 18:14 (UTC)

Use English, and you also:

JohnConnorFX commented on 2017-05-28 17:25 (UTC)

Agregar fakeroot como una dependencia antes de proceder a instalar google-chrome En la terminal: $ sudo pacman -S fakeroot Add fakeroot $ sudo pacman -S fakeroot

Det commented on 2017-05-25 07:50 (UTC)

commented on 2017-05-25 00:31 (UTC)

Please add fakeroot as dependency.

Skycoder42 commented on 2017-05-09 21:20 (UTC)

Please add libunity as optional dependency. It is required to show a progress when downloading files with KDE. Since Plasma 5.6 the unity launcher API is supported and thus works with KDE.

Det commented on 2017-04-24 15:10 (UTC) (edited on 2017-04-24 15:10 (UTC) by Det)


shsina commented on 2017-04-24 10:34 (UTC) (edited on 2017-04-24 10:36 (UTC) by shsina)

You should add xdg-utils to deps like chromium.

ppc52776 commented on 2017-04-21 01:54 (UTC)

if you have CompositorTileW problem, try to increase the font size by ctrl + '+' [ 1402.814324] CompositorTileW[6608]: segfault at 39 ip 00005609d8ce6af0 sp 00007f2ec1f19068 error 6 in chrome[5609d81ba000+6a81000] [ 1424.090013] CompositorTileW[6989]: segfault at 39 ip 00005609d8ce6af0 sp 00007f2ec1718068 error 6 in chrome[5609d81ba000+6a81000] [ 1494.299058] CompositorTileW[7361]: segfault at 39 ip 00005609d8ce6af0 sp 00007f2ec1718068 error 6 in chrome[5609d81ba000+6a81000] [ 1494.319679] CompositorTileW[7360]: segfault at 90 ip 00005609d8cf1c80 sp 00007f2ec1f19130 error 4 in chrome[5609d81ba000+6a81000] [ 1503.438023] CompositorTileW[7405]: segfault at 90 ip 00005609d8cf1c80 sp 00007f2ec1f191c0 error 4 in chrome[5609d81ba000+6a81000] [ 1782.250281] CompositorTileW[7798]: segfault at 39 ip 00005609d8ce6af0 sp 00007f2ec1717fd8 error 6 in chrome[5609d81ba000+6a81000] [ 1787.740639] CompositorTileW[7894]: segfault at 39 ip 00005609d8ce6af0 sp 00007f2ec1718068 error 6 in chrome[5609d81ba000+6a81000]

Det commented on 2017-04-07 12:10 (UTC)

I dunno? Rename your settings dirs (~/.cache/google-chrome/, ~/.config/google-chrome/), start in incognito, swap between the /usr/bin launcher and the one from /opt/google/chrome/ directly, extract the .deb and put the extension there, or try copying *all* the contents to /opt/google/chrome/, use "--enable-logging --v=1" (, etc.

DescartesHorse commented on 2017-04-07 11:48 (UTC)

@Det - yeah, it does work on other distro's like this. I dropped it in on a fresh Ubuntu with fresh Google Chrome installed, into /opt/google/chrome/extensions, and as soon as I opened up Chrome it appeared :( Went through that FAQ, and still in the same position. Any other ideas as to what may be going on, or how I can dig deeper to see where Chrome may be looking?

Det commented on 2017-03-31 07:03 (UTC) (edited on 2017-03-31 07:05 (UTC) by Det)

Easy way to check is for example to put an "exit" on line 51 and go to the pkg/opt/google/chrome/cron/ directory yourself (or Ctrl+C anytime after extraction and before that line): It basically seems to create and re-enable the Chrome repo for Debian/Ubuntu, which we don't need, since we don't have APT.

evilrob commented on 2017-03-31 03:09 (UTC)

Please excuse my noob question, but I'm trying to understand all these PKGBUILD files that I'm supposed to audit before letting them loose on my system. ;) What is the purpose of the Debian Cron jobs that Google has in their deb, and why are they removed in the PKGBUILD file?

Det commented on 2017-03-23 16:11 (UTC)

Should work like that, yes, dunno why not, because not a whole lot is changed in the packaging. Does it work on some other distro's Chrome? Maybe have a look at the "What are some common mistakes when installing with the preferences file?" in the FAQ?:

DescartesHorse commented on 2017-03-23 05:25 (UTC)

Hey, I've got a custom extension internally to our company (extension ID.json with update URL). Based on Google Chrome's documentation at I should be able to plop this in either /opt/google/chrome/extensions/ or /usr/share/google-chrome/extensions/ and have Google Chrome pick this up. This is not happening - wondering if this package is looking for extensions in another place? If so, where is that? :)

Det commented on 2017-03-19 10:23 (UTC)

@solus667, check first also, if it's updated.

Det commented on 2017-03-17 04:16 (UTC)

Anyone else who would not like to bother themselves with the Pinned Comments?

Musikolo commented on 2017-03-17 03:16 (UTC)

@Det, I'm getting the following validation error: [musikolo@MyPc google-chrome]$ makepkg -rs ==> Making package: google-chrome 57.0.2987.98-1 (jue mar 16 22:12:57 CDT 2017) ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Retrieving sources... -> Downloading google-chrome-stable_57.0.2987.98_amd64.deb... --2017-03-16 22:12:57-- Loaded CA certificate '/etc/ssl/certs/ca-certificates.crt' Resolving,,, ... Connecting to||:443... connected. HTTP request sent, awaiting response... 200 OK Length: 47312644 (45M) [application/x-debian-package] Saving to: ‘google-chrome-stable_57.0.2987.98_amd64.deb.part’ google-chrome-stable_57.0.2987.98_amd64.deb.part 100%[=========>] 45,12M 13,6MB/s in 3,4s 2017-03-16 22:13:01 (13,3 MB/s) - ‘google-chrome-stable_57.0.2987.98_amd64.deb.part’ saved [47312644/47312644] -> Found eula_text.html -> Found ==> Validating source files with md5sums... google-chrome-stable_57.0.2987.98_amd64.deb ... FAILED eula_text.html ... Skipped ... Passed ==> ERROR: One or more files did not pass the validity check! I've deleted the file and repeated the process a few times. Very weird the checksum doesn't match up! Could you please review to confirm what's wrong? Thank you!

guilherme.rialli commented on 2017-03-17 02:23 (UTC)

In the last update of the package had this error: Validating source files with md5sums ... Google-chrome-stable_57.0.2987.98_amd64.deb ... FAILED Eula_text.html ... Skipped ... Passed ==> ERROR: One or more files did not pass the validity check! ==> ERROR: Makepkg was unable to build google-chrome. Solution:$ makepkg -g >> PKGBUILD

lowsky commented on 2017-03-17 01:45 (UTC) (edited on 2017-03-17 01:46 (UTC) by lowsky)

Anyone having this issue? Validating source files with md5sums... google-chrome-stable_57.0.2987.98_amd64.deb ... FAILED eula_text.html ... Skipped ... Passed ==> ERROR: One or more files did not pass the validity check! ==> ERROR: Makepkg was unable to build google-chrome.

fuan_k commented on 2017-03-14 11:59 (UTC) (edited on 2017-03-14 12:08 (UTC) by fuan_k)

Anyone got an issue loading their user profile after latest update? :( Edit: nevermind, after starting chrome with --disable-extensions, then without again, it loaded fine again. Weird.

dare023 commented on 2017-03-13 07:24 (UTC)

After last update on every start I need to set it as a default browser, even if it is. Plasma5, tried to reset Chrome and confirmed it as a default browser.

fuan_k commented on 2017-03-01 22:55 (UTC)

@Det: yeah, I followed the updates on the bug report. Glad they managed to fix it. :)

Det commented on 2017-03-01 20:49 (UTC) (edited on 2017-03-01 20:50 (UTC) by Det)

"Updates: Status: Fixed Comment #14 on issue 678982 by Multiple GtkFileChooserDialog in parallel freeze Chrome" We, the People, have made the Chromium codebase Great Again. ( ͡° ͜ʖ ͡°)

zerophase commented on 2017-02-21 00:54 (UTC)

Don't believe so. Is chrome crashing for you?

yakshaver commented on 2017-02-21 00:47 (UTC)

Is anyone else getting a segfault on every first page load?

Det commented on 2017-02-03 20:11 (UTC)

Aaah, sorry, too careless. Fixed.

Xavion commented on 2017-02-03 19:44 (UTC)

A tiny bit of housekeeping: you've got an extra colon after 'kdialog' in the PKGBUILD (making it "kdialog::"). This prevents 'pacman' from displaying '[installed]' after it. I just went and tried to install 'kdialog', not realising I didn't need to.

Det commented on 2017-01-24 14:40 (UTC) (edited on 2017-01-24 14:40 (UTC) by Det)

> wiyarmir [1] added the following comment to google-chrome [2]: > > The build script complained about fakeroot command being missing although not being stated as a dependency (PKGBUILD) See:

Det commented on 2017-01-23 18:18 (UTC)

> DeadMetaler flagged google-chrome out-of-date on 2017-01-23 for the following reason: > > That's Chrome OS.

fuan_k commented on 2017-01-20 17:29 (UTC)

Thanks Det. Found that 55.0.2873.0 (unstable channel) was the version that started breaking. I've updated my bug report. Looks like someone is on the case now.

Det commented on 2017-01-19 15:11 (UTC)

> rborja flagged google-chrome out-of-date on 2017-01-19 for the following reason: > > I get a 404 error when running wget command. It's not out-of-date..

Det commented on 2017-01-17 18:48 (UTC) (edited on 2017-01-17 18:49 (UTC) by Det)

For sure, but I was just trying to help in bisecting. It's much easier for the devs when you've narrowed down the exact build ("1.14.1 works, 1.14.2 is broken").

fuan_k commented on 2017-01-17 18:11 (UTC)

Thanks Det, but I suspect it's an issue in the chromium code base as the bug didn't occur in 53.0.2785.143, and it occurs in 55.0.2883.87 (as stated in the bug report). Hopefully they will notice it and fix it eventually.

Det commented on 2017-01-15 17:46 (UTC)

By the way, you said "Did this work before? Yes 54. probably? Can't remember." in your report. You can find those older ones stacked in here: (link in google-chrome-dev PKGBUILD:

fuan_k commented on 2017-01-09 12:43 (UTC)

@mikioma totally unrelated, tested on different machines with different hardware (nvidia and amd). See bug report link for details. It's a regression in Chromium 55. Could be a library issue, or just the chromium code itself.

mikioma commented on 2017-01-09 08:39 (UTC)

@fuan_k If you happen to own an AMD/ATI card and Chrome's freeze status reverts by switching to a VT and swithing back to graphic desktop, then you're on the same page than other distros (Fedora among others). Seems to be related to graphics driver and get solved by launching Chrome with --disable-gpu flag... that makes it unusable, though (slow as pain...)

fuan_k commented on 2017-01-06 17:44 (UTC)

Can anyone reproduce the following bug (it will hang Chrome/Chromium) 1) set "Ask where to save each file before downloading" in settings 2) open TWO tabs in which downloads will start after a short countdown (for example: This works with any website, including video or image files when using "save as" very quickly in a row several times. 3) wait for the TWO GtkFileChooserDialog windows to open 4) click "cancel" in the top most "Save file as" dialog window (saving yields the same results in the end) 5) in the second (remaining) opened dialog, cancel as well (or save, same result next step anyway) 6) Chrome will start downloading the file if you chose to, but the user inputs are not registered anymore (mouse and keyboard alike). You will have to kill Chrome afterwards. I have submitted a bug report to Chromium bugtracker:

Det commented on 2016-12-29 17:45 (UTC) (edited on 2016-12-29 17:50 (UTC) by Det)

Stop sending you? :D I sent you _one_ _single_ email for the fact that you flagged this package twice because of your weird old build, when you can clearly see the package version up there is 55.0.2883.87. Any other emails would be due to the notifications *you* enabled. Btw. did you once 'ls' that directory or check the Wget output? I bet you get: $ ls google-chrome.tar.gz google-chrome.tar.gz.1 google-chrome.tar.gz.2 etc...

commented on 2016-12-28 23:01 (UTC)

home from work. here is a screenshot of my bash history from yesterday. you can see on entry 475 where i mashed tab for autocomplete. the other option was google-earth. i dunno what you want. i navigated to the aur package page, and i followed instructions about flagging the package when md5sums don't match. i appreciate the effort that you put into this and other packages. however, please stop sending me email.

storrgie commented on 2016-12-28 22:21 (UTC)

I use this package every day, thanks for the work you do @Det.

Det commented on 2016-12-28 16:37 (UTC)

You're not cd'ing to said directory, if you're fetching frigging google-chrome-stable_54.0.2840.90_amd64.deb.

commented on 2016-12-28 16:36 (UTC)

i navigate to the AUR package page, right click on download snapshot, get link url, paste into terminal, and wget. how could i have an old PKGBUILD?

Det commented on 2016-12-28 15:00 (UTC)

@blinkallthetime, stop flagging because of your goddamn 2 months old 54.0.2840.90 PKGBUILD. Clear your cache or manually fetch the package.

Det commented on 2016-12-24 18:38 (UTC)

I've never heard of that, but maybe this helps with your custom shell?

alien2050 commented on 2016-12-24 18:29 (UTC)

NOTE: I had ==> ERROR: PKGBUILD contains CRLF characters and cannot be sourced. I needed to dos2unix * beforehand to make it work.

jierumei commented on 2016-12-19 07:40 (UTC)

Yeah. If it is inconvenient, just ignore it. Thanks very much.

Det commented on 2016-12-18 06:07 (UTC)

Are you asking me, if I could recompile Google Chrome?

jierumei commented on 2016-12-18 01:25 (UTC) (edited on 2016-12-18 03:11 (UTC) by jierumei)

The fcitx input windows are not aligned with the target where it should be in HiDPI. It seems compile with recent gtk3 will solve it. Would you mind having a test?

Det commented on 2016-12-04 13:24 (UTC)

@zman0900: $ pacman -Qo /usr/lib/PepperFlash/ or: $ pacman -Ql google-chrome | grep flash

zman0900 commented on 2016-12-04 01:45 (UTC)

Why has the 'pepper-flash' provides been removed from this? The plugin is still in the package...

Det commented on 2016-12-02 18:34 (UTC)

Update incoming.

nlern commented on 2016-12-02 04:40 (UTC)

Please administrator remove the "Out-of-Date" Flag as the package is not. The current stable version of google-chrome is same as this package. Also the latest version given for flagging this package out-of-date is actually the current latest beta release and this package is stable release.

Det commented on 2016-11-28 10:31 (UTC) (edited on 2016-11-28 10:32 (UTC) by Det)

"blinkallthetime flagged google-chrome out-of-date on 2016-11-27 for the following reason: google-chrome-stable_54.0.2840.90_amd64.deb ... FAILED" This one is .100, my man. You've got an old PKGBUILD.

lonaowna commented on 2016-11-11 22:36 (UTC)

@Det, awesome. Thanks for maintaining!

Det commented on 2016-11-11 15:28 (UTC)

I dunno what's going on. Things happen and install files stay in place and stuff and it's a mystery!

lonaowna commented on 2016-11-11 13:14 (UTC)

Det, please read No single official package (take chromium as an example) is running things like update-desktop-database in the .install file.

Zillolo commented on 2016-11-11 12:07 (UTC)

Needs updating to 54.0.2840.100

jpprovost commented on 2016-11-10 09:10 (UTC)

Thank you for this package, it is really appreciated.

WoefulDerelict commented on 2016-10-21 16:51 (UTC)

Unfortunately the random segfaults persist on some systems running the recently promoted update to the stable 54.x family, 54.0.2840.71. I have not managed to sus out what changed after the 53.x series to introduce this behaviour as it does not seem to affect every system. It appears that Chrome has been sending out crash reports since I started running into this issue in 54.0.2840.59 so hopefully the team is not entirely unaware of this issue even tho I don't see a proper bug report for it.

Det commented on 2016-10-20 15:43 (UTC) (edited on 2016-10-20 17:32 (UTC) by Det)

53.x is available in many places: E: Does seem, btw, like there's a pattern with whenever Stable and Beta channels get bumped to be in sync there's always some segfault issues. It's not as prevalent when each channel's got its own major version.

Pandoras_Fox commented on 2016-10-20 15:42 (UTC) (edited on 2016-10-20 15:55 (UTC) by Pandoras_Fox)

Starting with Chrome 54.x, I'm getting some pretty regular segfaults. I tried going farther up the branches to beta/dev/chromium builds and they're still present there as well, so make sure you still have your 53.x package available to reinstall if this hits you. I'm still poking through the coredumps to try and figure this out. Sometimes it happens on idle, sometimes it happens while typing, but there doesn't appear to be an upstream bug report yet, so I'll work on getting that on their radar.

Det commented on 2016-10-18 15:58 (UTC)

The hell it is.

proof commented on 2016-10-18 14:35 (UTC)

Bad checksum, bad pkgver in build.

lonaowna commented on 2016-10-17 08:37 (UTC)

The update function in the .install file is no longer needed. It's function is already handled by /usr/share/libalpm/hooks/{update-desktop-database.hook,gtk-update-icon-cache.hook}

lsiudut commented on 2016-10-15 08:16 (UTC)

I can also confirm that 54.0.2840.59-1 is crashing very often. For me it happens mostly on complex sites like Facebook, eg. sending messages on chat is crashing browser immediately.

amrit commented on 2016-10-14 21:55 (UTC)

I am getting Segmentation fault too. Happens randomly every 2-10 minutes. It's really annoying.

WhatTheBuck commented on 2016-10-14 04:38 (UTC)

After updating, I have gotten several random segfaults. Program terminated with signal SIGSEGV, Segmentation fault. #0 0x0000562e00214583 in ?? () [Current thread is 1 (Thread 0x7f63bc24a700 (LWP 3025))] (gdb) bt #0 0x0000562e00214583 in () #1 0x00007f63bc249501 in () #2 0x0100000000000000 in () #3 0x000032827b536210 in () #4 0x0000000000000000 in () This has happened five times in the span of 4 hours.

keysona commented on 2016-10-11 02:28 (UTC)

@keysona ok, I have solved my problem. "A changes in your account that requires you sign in again." If you use gnome, you can go to Setting and add Google account in Online Account. From this:

keysona commented on 2016-10-11 02:02 (UTC)

Hi,guys. I met a problem. Every time I restart my computer, chrome will notify "A changes in your account that requires you sign in again."

Det commented on 2016-10-09 18:48 (UTC) (edited on 2016-10-09 18:49 (UTC) by Det)

"have done changes" :D. I'm sure you are aware of the "View Changes" link on the upper right.

j1simon commented on 2016-10-09 18:41 (UTC) (edited on 2016-10-09 18:43 (UTC) by j1simon)

You have done changes but you haven't updated the package version. How yaourt will update a package that has the same version?

Det commented on 2016-10-09 18:38 (UTC)

That's a fascinating email.

j1simon commented on 2016-10-09 18:36 (UTC)

Again a notification but there isn't update:

Det commented on 2016-10-09 14:30 (UTC)

The context menu should also be fixed?

Det commented on 2016-10-09 13:09 (UTC) (edited on 2016-10-09 13:10 (UTC) by Det)

I guess the cookie bug is this: (merged in 54.0.2840.50) Also, weird solutions:

agostonbarna commented on 2016-10-09 12:29 (UTC) (edited on 2016-10-09 12:30 (UTC) by agostonbarna)

@Det: On Gnome 3 the "New Incognito Window" context menu is missing. It can be enabled by appending the following lines to the google-chrome.desktop file: # [Desktop Entry] # ... Actions=new-window;new-private-window; [Desktop Action new-window] Name=New Window StartupWMClass=google-chrome Exec=/usr/bin/google-chrome-stable [Desktop Action new-private-window] Name=New Private Window StartupWMClass=google-chrome Exec=/usr/bin/google-chrome-stable --incognito

Det commented on 2016-10-06 15:23 (UTC)

I don't know how you received a "notification" about this?:

j1simon commented on 2016-10-06 15:21 (UTC)

Now I have received the notification about package update but the version is the same and yaourt doesn't update.

j1simon commented on 2016-10-04 16:02 (UTC)

Today I have "suffered" the cookie bug

jcarrus commented on 2016-10-04 15:51 (UTC)

@fusion809 I'm still experiencing the cookie bug in version 53.0.2785.143, are you experiencing this as well?

j1simon commented on 2016-10-03 15:05 (UTC) (edited on 2016-10-03 15:37 (UTC) by j1simon)

This error: I have to restart the browser with "chrome:restart"

Det commented on 2016-10-03 15:02 (UTC)

What is a "fail"? Crash? Freeze?

j1simon commented on 2016-10-03 15:00 (UTC)

With the last update, it fails often when I open several tabs. Anyone?

shug_org commented on 2016-10-01 19:21 (UTC)

Hi axedre, I had a similar problem, I had the desktop action in the ".desktop" file but didn't add those desktop actions to "Desktop Entry". Try adding "Actions=new-window;new-private-window;" make sure new-window and new-private-window, these can be whatever names you give your desktop actions. And have "Exec=/usr/bin/google-chrome-stable --incognito" for incognito desktop action. PS I am still pretty new at linux, so my apologies if I didn't use the right terminology.

axedre commented on 2016-09-24 19:03 (UTC)

Hi Det, good guess but it didn't work, as I feared: GNOME already wasn't among the targeted environments but the "New Window" command still showed it the left application tray, so adding "GNOME" to those two lines (the TargetEnvironment instruction in both the NewWindow and NewIncognito Shortcut Group sections) had no effect. It must be written somewhere else but I can't wrap my head around it...

Det commented on 2016-09-13 21:27 (UTC)

"Your package google-chrome [1] has been flagged out-of-date by flick36 [2]: Package is out of date" The hell it is.

Det commented on 2016-09-10 20:59 (UTC) (edited on 2016-09-10 21:00 (UTC) by Det)

I wonder, if changing the lines:  TargetEnvironment=Unity to:  TargetEnvironment=GNOME;Unity; in /usr/share/applications/google-chrome.desktop would accomplish that. You can type 'restart' or 'r' to reload the .desktop's:

axedre commented on 2016-09-10 18:55 (UTC)

Just wondering: how can I add a "New Incognito Window" to the context menu that pops up when I right-click on Chrome's icon in the favorites menu (gnome 3)? Googled a while but couldn't find a solution...

fusion809 commented on 2016-09-09 03:06 (UTC)

That cookie bug seems to be gone as of 53.0.2785.101.

Literallie commented on 2016-09-08 16:27 (UTC)

@fusion809: I'm experiencing that behaviour too.

Det commented on 2016-09-08 12:45 (UTC)


goxofy commented on 2016-09-08 04:49 (UTC)

hello! The package is out of date!

fusion809 commented on 2016-09-06 22:33 (UTC)

Nope I'm not working incognito and it seems I got it wrong, it's when I logout of Arch Linux / reboot, I loose my logins.

Det commented on 2016-09-06 14:27 (UTC)

You're losing the cookies. Incognito?

fusion809 commented on 2016-09-06 10:55 (UTC)

Am I the only one that is experiencing a bug where whenever I close Google Chrome all my login sessions (like for Facebook, Wikipedia, Twitter, GitHub, etc.) are lost and I have to login to each website again? This bug, I might add, is limited to Google Chrome. I logged into these sites in Chromium, then closed Chromium and re-opened Chromium and found that I'm still logged into all these sites.

technic93 commented on 2016-09-03 12:45 (UTC)

Hi, I think that postinstall updates are not necessary since pacman has its own hooks to update desktop database.

Det commented on 2016-09-02 20:51 (UTC)

So cool the interface in this new version...

WFV commented on 2016-09-02 05:05 (UTC) (edited on 2016-09-02 05:07 (UTC) by WFV)

This release wants password to unlock keyring whenever its launched. Is there way to disable/bypass the request?

Det commented on 2016-08-04 14:15 (UTC)

As mentioned, you can check the versions with a simple curl/gzip/awk flow. The full command (as well as the output) is in the same comment you quoted. If you've been using Arch that long, you should know better.

EisenSheng commented on 2016-08-04 13:56 (UTC)

> Check comment dates and content before posting. Wasn't able to build the package for like 2 days straight from direct pacaur -Sau run so I decided to look into here what's going on and post my experience since the package was marked out of date since yesterday and nothing interesting happened here. > I've been maintaining this package probably longer than you've been using Arch. To compare actual sizes, that's unlikely because I'm an arch user even since before chrome was a pipe dream in the head of some google techs. Less ego pls.

Det commented on 2016-08-04 12:49 (UTC) (edited on 2016-08-04 13:27 (UTC) by Det)

Check comment dates and content before posting. There's a command to check for updates right there, along with the output as of then. I've been maintaining this package probably longer than you've been using Arch. You can count on me to understand. Also, use Pastebin for logs.

EisenSheng commented on 2016-08-04 12:44 (UTC) (edited on 2016-08-04 13:49 (UTC) by EisenSheng)

> How so? > ✔ NOTE: If the md5sums don't match, don't post the makepkg output. Simply flag the package. I hope this rule doesn't match if the package maintainer doesn't understand why his package was marked as out of date. ;D

Det commented on 2016-08-02 14:37 (UTC) (edited on 2016-08-02 14:37 (UTC) by Det)

> Your package google-chrome [1] has been flagged out-of-date by hgaronfolo [2]: > Out of date. How so? $ curl -s | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }' stable: 52.0.2743.82 beta: 53.0.2785.34 unstable: 54.0.2810.2

Det commented on 2016-07-31 17:23 (UTC)


LuigiR0jas commented on 2016-07-31 17:20 (UTC)

I had the same font issue as @rthundr, and solved it the same way. gnome-themes-standard should really become dependencies for this package

grollinger commented on 2016-07-25 18:41 (UTC)

I just had an issue where all text on webpages that don't use webfonts of their own had disappeared. Running chrome from the console yielded an error message about a GTK Theme "adwaita" not being found. Installing gnome-themes-standard helped with that. Maybe this needs to be a dependency?

swanson commented on 2016-07-22 10:18 (UTC)

Netflix works fine here too but fonts and the Netflix gui are way too big. I shall look into DPI settings I think.

Zeroedout commented on 2016-07-21 22:05 (UTC)

@Det I just upgraded to Version 52.0.2743.82 (64-bit) and Netflix works fine... I seem to get as high quality video as I ever did.

dun commented on 2016-07-21 20:34 (UTC) (edited on 2016-07-22 11:45 (UTC) by dun)

Strangely one day after I've updated again and now it's running fine.

swanson commented on 2016-07-21 15:39 (UTC)

Font sizes in tabs and menus in Version 52.0.2743.82 (64-bit) are way too big!

zerophase commented on 2016-07-18 04:48 (UTC)

Is anyone having issues with Chrome forgetting computers that have logged into Google accounts with it?

rican-linux commented on 2016-05-20 22:40 (UTC)

I am noticing that youtube videos are playing but I am getting a blank screen.

Unia commented on 2016-05-15 19:32 (UTC)

Since pacman introduced hooks, running update-desktop-database from the .install file is no longer necessary.

johncgage commented on 2016-05-12 01:30 (UTC)

I ran the makepkg but everything worked ok. I'm using the same kernel version. It is associated with the sandbox. The only help I found was to use the --no-sandbox flag. But that seems an insecure route.

keiren commented on 2016-05-12 00:59 (UTC)

4.5.3 but I was having the same issue on 4.5.2

johncgage commented on 2016-05-12 00:39 (UTC)

What kernel version are you running?

keiren commented on 2016-05-12 00:28 (UTC)

I'm getting this when I try to run "google-chrome-stable" [1:1:0512/] Bad NaCl helper startup ack (0 bytes)

Det commented on 2016-04-29 06:59 (UTC)

Fixed? Updated? Yes, when I get home. Use 'SKIP' in the md5sums, or build with: $ makepkg (-si) --skipinteg

Svante commented on 2016-04-29 06:24 (UTC)

Will it be fixed in the near future? Need the stable version of Google Chrome.

Det commented on 2016-04-29 06:01 (UTC)

Because it "is out-of-date". Click the red link for the message.

Svante commented on 2016-04-29 05:32 (UTC)

Why is google-chrome "Flagged out-of-date"?

darkxsun commented on 2016-04-18 21:18 (UTC)

@mol3earth This should help you get started with the Arch User Repository:

mol3earth commented on 2016-04-18 20:02 (UTC)

Sorry if this is a total noob question. Coming to ArchBang from Debian based systems. But I can't seem to figure out how to install this thing, the Git Clone URL is a 404, and there is no tarball, or package to download... Does one copy and paste the PKGBUILD and run makepkg? Sorry and thanks.

silverhammermba commented on 2016-04-14 17:53 (UTC) (edited on 2016-04-14 17:54 (UTC) by silverhammermba)

@blanjp Can't you just edit the PKGBUILD to remove the EULA before you build? I doubt that Chrome actually needs it in order to run. You just need to delete three lines.

bianjp commented on 2016-04-14 12:46 (UTC) (edited on 2016-04-14 12:47 (UTC) by bianjp)

Could you please save the eula_text.html file into the package? In China, is blocked (while not), so downloading eula_text.html always failed, and I have to enable proxy and then rerun the update command.

Det commented on 2016-04-09 07:26 (UTC)

Nice tip. It's got them md5sums too.

z3ntu commented on 2016-04-07 19:38 (UTC)

FYI: $ curl | egrep "Package|Version" also displays the newest version (for google-chrome-stable, google-chrome-beta & google-chrome-unstable).

Det commented on 2016-04-07 17:58 (UTC)

Upstream problem.

morantron commented on 2016-04-07 17:57 (UTC)

This link is broken and the install no longer works

bcroq commented on 2016-03-29 10:18 (UTC)

unable to get the file upload dialog on twitter or github with 49.0.2623.110-1

michaelh commented on 2016-03-16 15:17 (UTC)

Hi Guys, Just upgraded from previous stable release to 49.0.2623.87-1 (had shockwave player issues) and now my printing from Chrome is borked. Running arch 4.4.1-2, cups 2.1.3-1, can't find anything wrong in my cups error log (Just keeps spitting back "No pages found!"). Having no problem printing from other programs, have also tried a range of printing methods (IPP,Jetdirect, LPD) etc. Cheers

Det commented on 2016-03-11 14:49 (UTC)

Please put all of it in Pastebin, and you should also use

resetko commented on 2016-03-10 21:16 (UTC) (edited on 2016-03-10 21:17 (UTC) by resetko)

After few last updates I've encountered problem with long shutdown/reboot time. After investigation I've found that problem comes only if I shutdown PC with chrome running. In shutdown log I've found timedout process in session-c2.scope: [ 423.510785] systemd[1]: session-c2.scope: Child 1337 belongs to session-c2.scope [ 451.565131] systemd[1]: systemd-logind.service: Got notification message from PID 486 (WATCHDOG=1) [ 452.065135] systemd-journald[207]: Sent WATCHDOG=1 notification. [ 452.065250] systemd[1]: systemd-journald.service: Got notification message from PID 207 (WATCHDOG=1) [ 511.565221] systemd[1]: session-c2.scope: Stopping timed out. Killing. [ 511.565398] systemd[1]: session-c2.scope changed stop-sigterm -> stop-sigkill [ 511.565563] systemd[1]: Received SIGCHLD from PID 945 (cat). [ 511.565582] systemd[1]: Child 945 (cat) died (code=killed, status=15/TERM) [ 511.565613] systemd[1]: session-c2.scope: Child 945 belongs to session-c2.scope And it was created by chrome, right? (raw systemd(1)-+-Thunar(692)-+-{gdbus}(863) ... |-xfce4-panel(690)-+-chrome(938)-+-cat(945) | | |-cat(946) | | |-chrome(989)-+-chrome(991) | | | |-{Chrome_ChildIOT}(993) | | | |-{SGI_video_sync}(1008) | | | |-{Watchdog}(992) | | | |-{WorkerPool/1005}(1005) | | | `-{handle-watcher-}(994) | | |-chrome-sandbox(948)---chrome(949)-+-chrome(954)-+-chrome(1001)-+-{Chrome_ChildIOT}(10+ | | | | | |-{CompositorTileW}(10+ | | | | | |-{CompositorTileW}(10+ | | | | | |-{CompositorTileW}(10+ | | | | | |-{CompositorTileW}(10+ ... This problem comes not always, but in ~85% of shutdowns with active chrome. How can I fix this? Can anyone confirm this?

Det commented on 2016-03-09 05:01 (UTC) (edited on 2019-04-05 02:27 (UTC) by Det)

✔ NOTE. If the sha512sums don't match, there's a new version. Simply flag the package.

You can check for new Linux releases in:, or use:

$ curl -sL | gzip -df | tr ' ' '\n' | grep -e name= -e ver= | cut -d '"' -f2 | sed 'N;s/\n/ /'

For new MD5: $ curl -sL | sed -n '35p'

Det commented on 2016-03-03 08:22 (UTC)

Kill it first.

justinnichols commented on 2016-03-03 00:40 (UTC)

After removing google-chrome manually and reinstalling via yaourt -S google-chrome, it acts as if it's obtained Google Chrome 49.0.2623.75, but after launching Chrome and viewing Help -> About Google Chrome, the version still shows 48.0.2564.109. Any thoughts?

Det commented on 2016-03-02 21:01 (UTC)

@xtrm0, "unmatching md5". No it isn't.

Det commented on 2016-03-02 20:53 (UTC) (edited on 2017-07-29 20:26 (UTC) by Det)

✔ OBS: 32-bit support for Google Chrome Stable Channel has been dropped as well. Either upgrade to 64-bit, or wait for phrakture to get hacked: -

commented on 2016-03-02 20:42 (UTC)

@Det, sorry. That was stupid of me.

Det commented on 2016-03-02 20:40 (UTC)

But please don't flag it _after_ it's updated.

Det commented on 2016-03-02 20:36 (UTC)

Read the description, please.

commented on 2016-03-02 20:35 (UTC)

Getting this- ==> Validating source files with md5sums... google-chrome-stable ... Passed eula_text.html ... Passed ==> Validating source_x86_64 files with md5sums... google-chrome-stable_48.0.2564.116_amd64.deb ... FAILED ==> ERROR: One or more files did not pass the validity check! ==> ERROR: Makepkg was unable to build google-chrome.

Det commented on 2016-02-10 06:54 (UTC) (edited on 2016-02-10 07:06 (UTC) by Det)

Oбновляются. Воспользуйтесь следующий раз Google Translate: Updated. Please Google Translate next time:

cyberlinux commented on 2016-02-09 22:33 (UTC)

Не совпадают md5sums Спасибо с Уважением Павел

cyberlinux commented on 2016-02-09 22:29 (UTC)

[cyberlinux@Lufix google-chrome]$ makepkg -sri ==> Сборка пакета: google-chrome 48.0.2564.103-1 (Ср фев 10 00:27:38 EET 2016) ==> Проверяю необходимые для запуска зависимости... ==> Проверяю необходимые для сборки зависимости... ==> Получение исходных файлов... -> Найден google-chrome-stable -> Найден eula_text.html -> Загрузка google-chrome-stable_48.0.2564.103_amd64.deb... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 45.8M 100 45.8M 0 0 1152k 0 0:00:40 0:00:40 --:--:-- 1065k ==> Проверка исходных файлов с помощью md5sums... google-chrome-stable ... Готово eula_text.html ... Готово ==> Проверка исходных файлов с помощью md5sums... google-chrome-stable_48.0.2564.103_amd64.deb ... СБОЙ ==> ОШИБКА: Один или более файлов не прошли проверку целостности!

UzielSilva commented on 2016-02-09 22:04 (UTC)

Hello. The md5sums don't match in: google-chrome-stable_48.0.2564.103_amd64.deb

Det commented on 2016-02-05 13:00 (UTC) (edited on 2016-02-05 13:01 (UTC) by Det)

Easiest way to test is always with a clean profile. Rename ~/.cache/google-chrome/ and ~/.config/google-chrome/.

commented on 2016-02-05 12:54 (UTC)

Anyone experiencing a total freeze when opening the Downloads page while downloading something?

Det commented on 2016-02-04 09:27 (UTC) (edited on 2016-02-04 09:29 (UTC) by Det)

Or, rather, unless you mean the data.tar.xz, the package type in Arch is .pkg.tar*, which by default is .pkg.tar.xz since 2010.

wchouser3 commented on 2016-02-04 08:45 (UTC)

Chrome does not update it's self for us. On .deb based package managers there's a cron job that checks for updates, but that just works for Ubuntu and Debian. There's another mechanism for Windows and OS X users, but again that has nothing to do with us. For us, Chrome has to be extracted from the .deb archive and converted over to the .xz archive in order to install.

Det commented on 2016-02-04 06:56 (UTC)

I'm not aware of Chrome automatically updating itself, besides on OS X, Windows, and Ubuntu (adds itself to /etc/apt/sources.list.d/).

eduncan911 commented on 2016-02-04 06:43 (UTC)

Here's a question: Does one need to keep updating this AUR package? Chrome self-updates to newer versions (when I installed it manually before). Or, does this package disable the auto-updating part of Chrome so that one must always keep this AUR package up to date?

Det commented on 2015-12-31 03:33 (UTC) (edited on 2016-01-23 20:52 (UTC) by Det)

That's good if they should be up-to-date as well, but there were enough people without one for me to just put the version number up there. In-and-of-itself, you don't do much with a PKGBUILD without pacman.

markzz commented on 2015-12-30 23:19 (UTC) (edited on 2015-12-30 23:21 (UTC) by markzz)

Any package belonging to the base (pacman is in base) or base-devel group are assumed to be installed and up to date on their systems before running makepkg, so adding pacman to the makedepends is redundant and unneeded.

Det commented on 2015-12-30 01:19 (UTC)

@markzz, "and the PKGBUILD doesn't appear to need any extra things from a specific version of pacman." Except it does actually quite clearly :D. "source_$CARCH" and "md5sums_$CARCH" were both features of makedepends=('pacman>=4.2.0').

markzz commented on 2015-12-29 23:20 (UTC) (edited on 2015-12-29 23:23 (UTC) by markzz)

Why is pacman a makedepends? It shouldn't be there because pacman is assumed to be installed on the system and the PKGBUILD doesn't appear to need any extra things from a specific version of pacman.

Det commented on 2015-12-24 16:37 (UTC)

The license prevents it:

tago commented on 2015-12-24 16:31 (UTC)

Why is not on official rep?

Det commented on 2015-12-12 11:14 (UTC)

Yeah, these are Google binaries and builds for Ubuntu/Debian, which I have no authority over whatsoever. But like the announcement says, the old ABI is still available.

catalin.hritcu commented on 2015-12-12 11:11 (UTC)

This package shows up on the list of packages that need to be rebuit after the recent switch to C++ dual ABI. Since this package seems directly based on Google binaries this will might need to be escalated with Google if that wasn't done already.

schultzter commented on 2015-11-05 02:58 (UTC)

So it looks like the problem I'm having is a known issue over at the Chromium project. The problem is some older CPU don't have the LAHF instruction.

schultzter commented on 2015-11-04 20:44 (UTC)

[abc@arch1 ~]$ google-chrome-stable Illegal instruction (core dumped) I had this same issue with google-chrome-beta and just started using stable. But I guess whatever Google did in -beta has made it's way into -stable now and I've got the issue here too. The issue still exists in -beta. I've tried running with the --debug flag but I don't get any additional output. Any ideas on how I can figure out what the issue is? Is there a dependancy I've got wrong? Is it my hardware or a driver? I have no idea. Any help would be appreciated.

Det commented on 2015-10-31 09:41 (UTC)

Rename ~/.config/google-chrome and ~/.cache/google-chrome, or have a look at the terminal output.

ricefan123 commented on 2015-10-31 09:39 (UTC)

Unable to run the application of current version

neTpK commented on 2015-10-13 19:19 (UTC)

NVIDIA: could not open the device file /dev/nvidia0 (Operation not permitted)

Det commented on 2015-10-12 08:05 (UTC) (edited on 2015-10-12 08:06 (UTC) by Det)

I mean, there never was a line like that. :) The Stable Channel doesn't include the channel name in the StartupWMClass anyway.

freijon commented on 2015-10-12 08:04 (UTC)

without the "- " in front, it was meant to indicate that this line should be replaced.

Det commented on 2015-10-07 17:43 (UTC)

There is no - sed -i "/Exec=/i\StartupWMClass=Google-chrome-$_channel" "$pkgdir"/usr/share/applications/google-chrome-$_channel.desktop

freijon commented on 2015-10-07 16:46 (UTC)

Please modify the following line in the PKGBUILD: - sed -i "/Exec=/i\StartupWMClass=Google-chrome-$_channel" "$pkgdir"/usr/share/applications/google-chrome-$_channel.desktop + sed -i "/Exec=/i\StartupWMClass=google-chrome-$_channel" "$pkgdir"/usr/share/applications/google-chrome-$_channel.desktop The upper case "G" causes plank to show double icons when Chrome is running

Det commented on 2015-09-25 05:21 (UTC) (edited on 2015-10-03 08:02 (UTC) by Det)

@jeffska, you're 2 days late.

jeffska commented on 2015-09-25 01:58 (UTC)

@Det: Why flagged? Because: ==> Validating source files with md5sums... google-chrome-stable_45.0.2454.99_amd64.deb ... FAILED

Det commented on 2015-09-23 17:57 (UTC)

@elemongw, why?

Det commented on 2015-09-09 23:29 (UTC)

Yepyepyep. YepyepyepYepyepyepYepyepyepYepyepyep. Yepyep.

m45t3r commented on 2015-09-09 12:24 (UTC)

@BrtH Yep, according to Docky wiki, the WMClass in the current .desktop file is wrong: Running "xprop | grep WM_CLASS | awk '{print $4}'" and clicking in a Google Chrome window results in: "Google-chrome" So the entry in the .desktop file should be "StartupWMClass=Google-chrome".

BrtH commented on 2015-09-07 16:23 (UTC)

The icon is displayed twice on the Plank dock. This is solved by removing the '-stable' from 'StartupWMClass=Google-chrome-stable' in google-chrome.desktop. This problem can be solved by removing '-$channel' on line 53 of PKGBUILD. I, however, do not know what the consequences can be on other machines than my own or when used with a different dock than Plank.

ninian commented on 2015-09-04 19:05 (UTC)


Det commented on 2015-09-04 19:00 (UTC)

I mean "no" to the "no". Link the thread next time: Didn't bump the pkgrel. It's quicker to add the quotes yourself.

commented on 2015-09-04 18:40 (UTC)

Det commented on 2015-09-04 18:28 (UTC)


ninian commented on 2015-09-03 22:25 (UTC)

Methinks the last line of the google-chrome-stable script should rather be: exec /opt/google/chrome/google-chrome $CHROME_USER_FLAGS "$@" That is, with the $@ double-quoted?

Joan31 commented on 2015-09-02 11:50 (UTC)

it's ok now with google-chrome-stable_45.0.2454.85_amd64.deb thanks

Joan31 commented on 2015-09-01 20:31 (UTC)

Hello the md5sums don't match ==> Validation des fichiers sources avec md5sums... google-chrome-stable_44.0.2403.157_amd64.deb ... ÉCHEC (FAIL)

Det commented on 2015-08-30 06:06 (UTC)

Lol. Well. I don't have 'pacman' (which, if you don't have, installing packages is kind of pointless anyway), but I have "pacman 4.2.0 or later", which you do need, if you want to build this package.

lybin commented on 2015-08-28 08:08 (UTC)

@Det, You have 'pacman' in 'makedepends'. pacman is included to base-devel group. Note: The group base-devel is assumed to be already installed when building with makepkg. Members of this group should not be included in makedepends array.

Det commented on 2015-08-28 04:51 (UTC)


lybin commented on 2015-08-28 04:49 (UTC) Note: The group base-devel is assumed to be already installed when building with makepkg. Members of this group should not be included in makedepends array. local/pacman 4.2.1-2 (base base-devel)

Det commented on 2015-08-26 17:43 (UTC)

You could have set that anywhere else too, but there's now support for flags read directly from ~/.config/chrome-flags.conf (no defining CHROM*_USER_FLAGS variables). No bump in pkgrel. Should work.

lybin commented on 2015-08-23 09:25 (UTC)

If someone has problem with DPI, need to add first line CHROMIUM_USER_FLAGS="--force-device-scale-factor=1" to /opt/google/chrome/google-chrome

Det commented on 2015-08-21 01:46 (UTC)

It's .157.

T1000 commented on 2015-08-20 20:32 (UTC)

Version: 44.0.2403.155-1 md5sums_x86_64 checksum needs to be changed to 24097e63c43976e9f9304edd5bcd4ac0

Det commented on 2015-08-14 23:28 (UTC)

@hewr1993, why..?

Det commented on 2015-08-12 13:11 (UTC)

Unfortunately, 1) That format won't work, because you need the build number 2) It means you need to generate the link, making it harder to update it yourself 3) People will still post makepkg outputs when the build fails.

delx commented on 2015-08-12 09:53 (UTC)

Hi, thanks for maintaining this package! Would you consider modifying it to use versioned URLs? This way if Google pushes an update it will fail with a 404 instead of a validity check :) source_i686=("google-chrome-${_channel}_${pkgver}_amd64.deb::${_channel}_${pkgver}-1_i386.deb") source_x86_64=("google-chrome-${_channel}_${pkgver}_amd64.deb::${_channel}_${pkgver}-1_amd64.deb")

Det commented on 2015-08-11 18:09 (UTC)

That's the new version coming up, changing the checksums. There's almost every release somebody posting the makepkg output of the failed validity check. You won't find it in Google, if you include the version number.

j3sk0la commented on 2015-08-11 18:06 (UTC)

Working now. Thanks. Feel free to PM me what was lol and I will avoid repeating in the future.

Det commented on 2015-08-11 18:02 (UTC)


j3sk0la commented on 2015-08-11 17:58 (UTC)

Hi, I get the following error when trying to install or update this package. I have updated my system (Syyu, Syu, and Syua, Syyua), refreshed my keyrings, restarted my system, and removed ~/.config/google*, ~/.cache/google*, and /tmp/*. I checked the forums but didn't see anything. Yaourt output: ==> Validating source files with md5sums... google-chrome-stable_44.0.2403.130_amd64.deb ... FAILED ==> ERROR: One or more files did not pass the validity check! ==> ERROR: Makepkg was unable to build google-chrome. Please advise. Thank you in advance.

gideonite commented on 2015-07-30 15:29 (UTC)

Sorry for taking so long to respond @dat. I tried running with a clean profile by removing ~/.config/google-chrome and success! I wonder what was causing the problem. Now my window manager (i3) is also registering some sort of chrome daemon in the status bar. Thanks for the help.

Det commented on 2015-07-24 18:38 (UTC)

I just tried launching .107 a couple of times, and it works just fine so far.

Det commented on 2015-07-24 18:36 (UTC)

You mean with 44.0.2403.107 (update script running now) or .89? The 43.x on is available at (or /var/cache/pacman/pkg/google-chrome-*): Does downgrading to that or running with a clean profile (~/.config/google-chrome/, ~/.cache/google-chrome/) work?

gideonite commented on 2015-07-24 18:31 (UTC)

I'm getting some very strange behavior since my last update. I have trouble starting google-chrome-stable --- it will randomly fail. Finally, after trying a number of times, it sort of "catches on" and is fine after that. When I run strace the program halts consistently at a read command. Here are the last few lines of strace's output: Also worth mentioning is that each time I run google-chrome-stable and no browser window shows up, chrome processes are kicked off (I can see them using `ps aux`). Any ideas?

Paviluf commented on 2015-07-15 00:17 (UTC)

It's the first time I try to install Google Chrome and I misunderstood your previous comment. I thought you won't update it anymore. Sorry ;)

Det commented on 2015-07-15 00:12 (UTC)

No.. I've had the habit of keeping this up-to-date since 2013, you just have to give me more than 4 hours, when updating is impossible.

Paviluf commented on 2015-07-15 00:04 (UTC)

Hello Det, Do you plan to keep the PKGBUILD up to date after each Google Chrome release or do we need to update Chrome manually ? Thank you.

Det commented on 2015-07-14 21:29 (UTC)

Absolutely no need, when you can just do: $ makepkg -si --skipinteg

toppy commented on 2015-07-14 21:29 (UTC)

Edit PKGBUILD and replace md5 checksums with the following ... (generated from 'makepkg -g') md5sums=('b7e752f549b215ac77f284b6486794b6') md5sums_i686=('9a806fa944c59ce398b13c8c79bca123') md5sums_x86_64=('b26558f6be7834d36a58e70348082203') Package builds ok for me

Det commented on 2015-07-14 20:13 (UTC)

@tmoore, 1) I know, this was flagged (see top of PKGBUILD), and 2) see:

commented on 2015-07-14 14:06 (UTC)

Updates to v. 43.0.2357.134

darkxsun commented on 2015-06-22 16:02 (UTC)

Flagging out-of-date is good enough...

CurtisLeeBolin commented on 2015-06-22 15:45 (UTC)

$ md5sum google-chrome-stable_43.0.2357.125_amd64.deb b02492c9eb6131bb0a43cb8a906bd6e5 google-chrome-stable_43.0.2357.125_amd64.deb

catalin.hritcu commented on 2015-06-20 19:48 (UTC)

Even with .125-1, I still need to pass --force-device-scale-factor=1 for things not to be horribly scaled.

puppetluva commented on 2015-06-11 17:40 (UTC)

Flagged out of date: google-chrome-stable_43.0.2357.124_amd64.deb ... FAILED ==> ERROR: One or more files did not pass the validity check!

Det commented on 2015-05-27 16:19 (UTC)

Make sure you kill the processes first.

costis commented on 2015-05-27 16:18 (UTC)

Strange thing happens. From the command line I get the correct chrome version, but from "Settings > About Google Chrome" in the browser my previous, 42.* version. google-chrome-stable --version Google Chrome 43.0.2357.81

Det commented on 2015-05-26 05:45 (UTC)

@sli908, but that's awesome that they have a bug tracker for that kinda thing ^^:

Det commented on 2015-05-24 05:22 (UTC)


nagar commented on 2015-05-23 23:16 (UTC)

when start google chrome, it has no icon on task bar, task bar icon is missing but app icon exist, Thanks

Det commented on 2015-05-23 17:07 (UTC)

Of COURSE it is, stop SAYING it!

marmotz commented on 2015-05-23 14:12 (UTC)

@sli908 @isra17 same issue here with MATE

sli908 commented on 2015-05-22 01:59 (UTC)

@isra17 I checked and Chromium seems to be having the same issue.

isra17 commented on 2015-05-21 23:26 (UTC)

@sli908 I have the same issue using XFCE. Chrome windows now have the generic icon in the bar.

sinatosk commented on 2015-05-21 07:40 (UTC)

@sandnon thanks that works for me

oryxfea commented on 2015-05-21 07:19 (UTC)

@petterk @sinatosk @catalin.hritcu I think I found a temporary solution. Just add `--force-device-scale-factor=1` to command line. This issue may relate to HiDPI.

sli908 commented on 2015-05-21 00:51 (UTC)

I don't have the rendering issues either. I'm getting the "generic window" icon in MATE as well. I tried a fix I saw below that suggested removing the Version line from the .desktop file, and that did not work.

VDP76 commented on 2015-05-20 23:29 (UTC)

I do not seem to be affected by the rendering issues. However now chrome icon in tint2 taskbar is not displayed any more, I see the "generic window" icon instead. Changing icon set or regenerating the icon cache does not help. Is anyone having this issue too?

neTpK commented on 2015-05-20 15:54 (UTC)

So its not only me. My issues is its abit larger than before, but the fonts are so unsharp and "foggy". Hard to read many websites. Zooming out the page does not help.

catalin.hritcu commented on 2015-05-20 14:54 (UTC)

The font scaling problem only affects Chrome v43. Downgrading to v42 solved the problem for me, temporarily at least since I would like to eventually upgrade again.

catalin.hritcu commented on 2015-05-20 14:48 (UTC)

@petterk, @sinatosk: Same font scaling problems here with 1920x1080 on MATE at 150dpi. It affects both the rendered page and Chrome itself and various page elements are affected to different degrees (GMail is a mess, the text in this text field is much larger than the other fonts on the page, etc.)

sinatosk commented on 2015-05-20 14:07 (UTC)

@petterk everything is larger for me... so large ( first display 1920x1080 ) the settings menu shows on my second display ( 2560x1600 ) the font looks like normal to medium when set to very small or small Google changed something I guess...

neTpK commented on 2015-05-20 12:00 (UTC)

Anyone else have weird looking fonts after the last update?

Det commented on 2015-05-20 06:43 (UTC)

Flag it.

samehkamaleldin commented on 2015-05-20 06:42 (UTC)

==> Validating source files with md5sums... google-chrome-unstable_44.0.2398.0_amd64.deb ... FAILED ==> ERROR: One or more files did not pass the validity check! ==> ERROR: Makepkg was unable to build google-chrome-dev.

Det commented on 2015-05-19 19:43 (UTC)

Also, as long as you don't clear your cache, they can be found in: /var/cache/pacman/pkg/.

madscience commented on 2015-05-19 19:41 (UTC)

also here: definitely an issue with upstream.

Det commented on 2015-05-19 19:18 (UTC)

Yeah, they're all available in: See, if that helps.

madscience commented on 2015-05-19 19:15 (UTC)

Does anyone have an older .deb kicking around? Google only has stable-current on the server, and I'd like to temporarily downgrade.

brittyazel commented on 2015-05-19 19:07 (UTC)

So version 42 worked flawlessly with my touchscreen, and all touch events were automatically enabled, but with 43 (and prior to 42) I am back to touch functionality not being enabled unless chrome is launched with touch. What happened lol?

madscience commented on 2015-05-19 18:49 (UTC)

If it helps, the .desktop file shows the correct icon, it looks fine in the menu and on the panel. It's just the taskbar icon that doesn't pick up.

madscience commented on 2015-05-19 18:41 (UTC)

That didn't work, unfortunately. I commented out the Version=1.0 line, ran those two commands and then also tried logout/login for good measure.

Det commented on 2015-05-19 18:26 (UTC)

Remove the "Version" line from the .desktop (/usr/share/applications/google-chrome.desktop).

madscience commented on 2015-05-19 18:25 (UTC)

I didn't see any suggestion in the link to the git commit...?

Det commented on 2015-05-19 18:21 (UTC)

No, but did you actually do what was suggested in the link?

madscience commented on 2015-05-19 18:18 (UTC)

ran both of those as root, logged out and back in... same issue.

Det commented on 2015-05-19 18:11 (UTC)

Does that help?: Run "update-desktop-database" and maube "xdg-icon-resource forceupdate --theme hicolor" as root afterwards.

madscience commented on 2015-05-19 18:09 (UTC)

Since the latest update, the taskbar icon is missing (xfce). I'm using numix circle icon theme, but I've tried default gnome theme and hi-color, but all I get is the generic icon.

Det commented on 2015-05-12 18:27 (UTC)

Wow. 1000 votes:

Krieghof commented on 2015-04-16 18:00 (UTC)

@Det Thank you very much. My / partition was full. Now it works flawless.

Det commented on 2015-04-16 17:14 (UTC)

Nice desktop. I would try rebooting or with a clean profile first. Googling doesn't reveal much. Gdb'ing, fsck'ing, or checking if your partitions have free space might help:

Krieghof commented on 2015-04-16 16:42 (UTC)

Can't start chrome after the latest update. The Terminal says "Bus error (core dumped)"

oryband commented on 2015-04-09 11:53 (UTC)

@sugatang.itlog - if you have scroll and mouse pointer issues, check the following chromium issue: They're talking about a temporary fix, take a look.

ekkelett commented on 2015-04-08 05:47 (UTC)

The issue reported by martinpoljak seems to have been fixed, as it worked fine here. Anything amiss in regards to Widevine?

martinpoljak commented on 2015-04-06 07:22 (UTC)

-> Downloading google-chrome-stable_41.0.2272.118_amd64.deb... Initializing download: HTTP/1.0 404 Not Found Something is wrong.

sugatang.itlog commented on 2015-04-05 06:08 (UTC)

hi, have problems with the new google-chrome (google-chrome 41.0.2272.118-1). I cannot highlight text, mouse pointer is not changing to proper pointer when hovering to a link. Though no problem in firefox. Thanks!

nick commented on 2015-04-01 22:18 (UTC)

new md5sums: ef61d7c2b66ba26b1f297f82e7e1a52b google-chrome-stable_current_amd64.deb 3d9b5f96dbdeaca079b39d41135b662b google-chrome-stable_current_i386.deb

fichtner commented on 2015-04-01 18:56 (UTC)

md5sums currently out of date. Was able to install with modifying PKGBUILD

zman0900 commented on 2015-03-20 04:29 (UTC)

└──╼ curl -s | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }' stable: 41.0.2272.101 beta: 42.0.2311.50 unstable: 43.0.2327.5

Det commented on 2015-03-20 01:37 (UTC)

Still .89: 0 ✓ det_000@Awesomeness-PC /cygdrive/c/Users/det_000/Desktop $ curl -s | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }' stable: 41.0.2272.89 beta: 42.0.2311.50 unstable: 43.0.2327.5

Det commented on 2015-03-15 02:46 (UTC)

What types of files are you trying to open that don't show Chrome in the menu? E: I'm not gonna put opening files through Incognito the default.

dejavu commented on 2015-03-12 21:59 (UTC)

Would it be possible to change the google-chrome.desktop file (for instance with sed) and use following lines in it?! The first line would allow selecting the application in "Open With" in the file managers. Exec=/usr/bin/google-chrome-stable --incognito %U TryExec=/usr/bin/google-chrome-stable

JetJaguarXP commented on 2015-03-11 00:08 (UTC)

Thank you Det!

Det commented on 2015-03-11 00:03 (UTC)

Stop, please.

Det commented on 2015-03-11 00:02 (UTC)

Well, you don't have to flag it _anymore_, I already updated it.

Det commented on 2015-03-10 23:27 (UTC)

Yep, it means there's a new version so the md5sum doesn't match anymore. You can use the *Flag package out-of-date* button in the future.

JetJaguarXP commented on 2015-03-10 23:20 (UTC)

Anyone else getting this? $ makepkg ==> Making package: google-chrome 41.0.2272.76-1 (Tue Mar 10 16:16:04 MST 2015) ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Retrieving sources... -> Found eula_text.html -> Downloading google-chrome-stable_41.0.2272.76_amd64.deb... ==> Validating source files with md5sums... google-chrome-stable_41.0.2272.76_amd64.deb ... FAILED ==> ERROR: One or more files did not pass the validity check!

abhisit commented on 2015-03-05 04:33 (UTC)

Thanks a lot.

cellisten commented on 2015-03-03 19:58 (UTC)

New md5sum and also changed from data.tar.lzma to data.tar.xz

Det commented on 2015-03-01 16:00 (UTC)

@shibumi, why?

brittyazel commented on 2015-02-28 09:35 (UTC)

And I disable all extensions :-/

brittyazel commented on 2015-02-28 09:35 (UTC)

No I dont actually. I made sure to disable all the things that I suspected would keep a zombie process alive.

Det commented on 2015-02-28 08:51 (UTC)

It might be some extension or app that's causing it. Do you have "menu → Settings → Continue running background apps when Google Chrome is closed" on (default)?

brittyazel commented on 2015-02-27 23:32 (UTC)

Ahh I see that. Unfortunately that didn't help. I think I know what is happening though. Occasionally when closing chrome, a process will linger. Not every time, but about 1/3 times. If I open a new chrome instance before that previous process is closed, I throw database profile errors, as the database is still locked as long as that lingering process remains open. I have found that ending all the chrome processes one by one will allow me to open chrome without an issue, and there is always one that is stubborn to close and forces me to kill the process. As to why a process lingers open, I have no idea.

Det commented on 2015-02-27 23:20 (UTC)

You know, "menu → Settings → Manage search engines...", which are saved in the 'Web Data' database.

brittyazel commented on 2015-02-27 23:18 (UTC)

What do you mean by one of my search engines?

Det commented on 2015-02-27 23:08 (UTC)

If it's on four computers with different settings, then that's systemic. Maybe some extension is screwing it up, or one of your search engines has a corrupt configuration, which is getting synced to all the rest of them.

brittyazel commented on 2015-02-27 23:06 (UTC)

Nvm, that was not the fix to my problem. If I close and reopen chrome in too close of a succession I get a "profile could not be loaded" error message. Looking in system-monitor shows a chrome process remaining open even though all windows are close (and I have the stay open option turned off). I think it has something to do with a local sql database, as deleting 'web-data' in the /config/google-chrome/default will fix the issue temporarily. And it happens on four of my computers with completely different architecture.

Det commented on 2015-02-27 23:01 (UTC)

Shouldn't you clear your cache then?

brittyazel commented on 2015-02-27 23:00 (UTC)

I kept having my profile fail to load nearly every time I opened Chrome, however appending "--disk-cache-dir=/tmp/cache" to the end of the exec lines in the .desktop file has helped tremendously. Is there any way to do this globally so that way I do not have to redo it each update?

Det commented on 2015-02-22 16:26 (UTC)

AJSlye, reason?

WorMzy commented on 2015-02-19 16:50 (UTC)

40.0.2214.115 just out:

Det commented on 2015-02-12 07:13 (UTC)

It's because of: warning: directory permissions differ on /usr/share/doc/google-chrome-stable/ filesystem: 700 package: 755 Since it's a directory, the permission can't be automatically changed by the: msg2 "Fixing permissions of documentation folder..." chmod 755 "$pkgdir"/usr/share/doc/google-chrome-$_channel/

Moviuro commented on 2015-02-12 07:07 (UTC)

Hi there, Looks like the following folder has wrong permissions: drwx------ 1 root root 24 Jan 29 11:35 /usr/share/doc/google-chrome-stable Anyway, thanks for the package ;)

Det commented on 2015-02-11 13:30 (UTC)

@mgb, that's for Chrome OS:

Det commented on 2015-02-07 17:10 (UTC)

This is true.

weirddan455 commented on 2015-02-07 17:01 (UTC)

We're getting the checksum from you though, not from Google, and you have nothing from Google to verify the file against. If someone were to replace the file on Google's server with a malicous one (which, yes, is unlikely but not impossible. Google could be hit with a 0 day vaunerability just like anyone else) then you would update the checksum on the PKGBUILD thinking the file was legitamite. That only needs one point of entry. If Google were using PGP signautures then that would require two points of entry. The attacker would have to both replace the file AND steal Google's private key to sign the file.

Det commented on 2015-02-07 16:36 (UTC)

@weirddan455, that's assuming, of course, that you'd be able to break into two distinct places. The only practical way to do Google is to be walking into the server room, hogging a few, and coming up with a great excuse. @ilpianista, are you using Manjaro, or did you touch the line below "Symlinking missing Udev lib..."?

ilpianista commented on 2015-02-07 16:26 (UTC)

/opt/google/chrome/chrome: error while loading shared libraries: cannot open shared object file: No such file or directory

weirddan455 commented on 2015-02-07 14:28 (UTC)

Any checksum, no matter the algorithm, won't protect you from someone malicously altering the file. Reason being is that if someone manages to replace the file you're downloading it probably wouldn't be too hard for them to simply replace the checksum as well to match the malicous file. The only thing the checksum is good for the way we use it here is to protect against accidental data corruption during the download. MD5 will work just fine for that. Hell even CRC would probably do the job. If you want real data integrety protection, we really PGP signatures. Sadly I don't think Google signs their Chrome DEBs though.

Det commented on 2015-02-07 05:38 (UTC)

We can learn something new every day, but now that the aftermath of my controversial and time-consuming edit[1] for our awesome Wiki admin, Kynikos, has been settled[2], I've now explained the differences between collision and preimage vulnerabilities in MD5, SHA-1 and SHA-2 in the PKGBUILD article: This took quite a bit of researching, but it basically means that having MD5 checksums for file integrity are actually fine for now. You don't want to use it for SSL certificates, but carrying out a successful preimage attack requires, even in theory, approximately 2^123.4 ≈ 1.4 × 10^37 operations, which is completely impractical. [1] = [2] =

tormund commented on 2015-02-06 18:10 (UTC)

Latest release won't let you press enter to search terms in the omnibox, but it will let you go to suggested websites.

Det commented on 2015-02-05 17:25 (UTC)

You need to press "Flag package out-of-date" from the right :).

Xiaoming94 commented on 2015-02-05 17:23 (UTC)

plz update this package

Det commented on 2015-02-05 10:11 (UTC)

With this specific package it doesn't really make much of a difference (I'd like to see anybody breaking Google's servers), but I could actually convert all of my packages to SHA-256. There's barely any performance impact: $ time md5sum google-chrome-stable_current_amd64.deb | tail -1 md5sum google-chrome-stable_current_amd64.deb 0.11s user 0.01s system 105% cpu 0.118 total $ time sha256sum google-chrome-stable_current_amd64.deb | tail -1 sha256sum google-chrome-stable_current_amd64.deb 0.27s user 0.03s system 96% cpu 0.308 total The poor algorithm has indeed been tortured ever since 2004 (, and there's no real reason to still be using it. Everybody's just too lazy to do anything about it. Even the Wiki still mentions md5sums:

vikstrous commented on 2015-02-05 06:16 (UTC)

Please use something better than md5, like sha256.

Det commented on 2015-02-04 13:50 (UTC)

This wasn't really all that hard to find: It's got to do with libgtk indeed.

gauravpadia commented on 2015-02-04 13:44 (UTC)

I am also experience the same issue as specified by @radikal. Chrome randomly crashes when saving image or any page. I am using awesome window manager. Here is the dmesg log: [ 8034.745932] chrome[1112]: segfault at 30 ip 00007ffd770f0186 sp 00007fff0360d170 error 4 in[7ffd76e31000+436000] [12221.628406] chrome[5593]: segfault at 30 ip 00007f73be130186 sp 00007fff75c3e9a0 error 4 in[7f73bde71000+436000]

Det commented on 2015-02-04 00:49 (UTC)

You can use: $ curl -s | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }' beta: 41.0.2272.35 stable: 40.0.2214.95 unstable: 42.0.2292.0

skogler commented on 2015-02-04 00:16 (UTC)

checksum seems to be incorrect, for me (64 bit) the correct md5sum is 361aea21e6e3acaa805af17edb4ffe49

Det commented on 2015-01-29 22:02 (UTC)


commented on 2015-01-29 21:49 (UTC)

The .desktop file for me has the wrong StartupWMClass. It gets set to Eula and then Plank doesn't recognize open windows of it. StartupWMClass=Eula

silverhammermba commented on 2015-01-28 15:31 (UTC)

I also get a segfault when saving files, but if I just reopen Chrome and my old tabs (Ctrl+Shift+T) trying to save a second time works.

Det commented on 2015-01-27 14:32 (UTC)

If you guys get that with a clean profile for both the Beta and Dev Channel as well, then you should report a bug.

radikal commented on 2015-01-27 14:27 (UTC)

Hi Catalin, That said I really dont know if this is an chrome issue or libgtk? Ive tried to googlefu the issue but failed. My collegue has the same problem, both of us are using arch and awesome as windowmanager. Dont know any other way to reproduce than using "Save page as". Every time I try I see the save window starts drawing and then *poff* chrome dies and this error message in dmesg comes up.

catalin.hritcu commented on 2015-01-27 14:23 (UTC)

@radikal: Tried "Save page as" and it worked fine. Anything special I need to do to reproduce this?

radikal commented on 2015-01-27 14:20 (UTC)

Am I the only one seeing crashes when using "Save page as" [ 5924.345719] chrome[625]: segfault at 30 ip 00007f8fd91a3956 sp 00007fff586e2a00 error 4 in[7f8fd8ee5000+435000] [172071.870967] chrome[3162]: segfault at 30 ip 00007f077a445956 sp 00007fff00e16fe0 error 4 in[7f077a187000+435000] [243127.630583] chrome[28297]: segfault at 30 ip 00007f088b43f956 sp 00007fff59e99f90 error 4 in[7f088b181000+435000] [607124.690762] chrome[24993]: segfault at 30 ip 00007f6ce0706956 sp 00007fffd6963c30 error 4 in[7f6ce0448000+435000]

Det commented on 2015-01-27 13:04 (UTC)

@bmbaker, @catalin.hritcu, @LarcBP, yes guys, when a new version's up, the md5sums won't match anymore. We actually don't need any more people confirming this problem each time a new version's up. You can consider it confirmed. Also, I know you're eager and everything, but please, just let me update it doesn't have to be removed...:

LarcBP commented on 2015-01-27 12:06 (UTC)

to solve this last problem, change the line on PKGBUILD: md5sums_x86_64=('0bd66eeac16d62778ad7cb33d608eb68') to md5sums_x86_64=('ac0e961f5001a38c087942570ac65178')

catalin.hritcu commented on 2015-01-27 09:20 (UTC)

I can confirm bmbaker's problem. I get a similar problem with aura: ==> Validating source files with md5sums... google-chrome-stable_40.0.2214.91_amd64.deb ... FAILED

bmbaker commented on 2015-01-27 08:44 (UTC)

there seems to be a problem with PKBUILD Validating source files with md5sums... google-chrome-stable_40.0.2214.91_amd64.deb ... FAILED ==> ERROR: One or more files did not pass the validity check! ==> ERROR: Makepkg was unable to build google-chrome.

Det commented on 2015-01-16 09:29 (UTC)

Pacman 4.2.0 is apparently in [testing] for Manjaro: You can get the previous (4.1.x compatible) PKGBUILD in:

annhill commented on 2015-01-16 09:24 (UTC)

If I build this in Manjaro (0.8.11), I get the following error: error: target not found: pacman>= 4.2.0 My version of pacman is 4.1.2, how can I fix this?

nogweii commented on 2015-01-13 21:13 (UTC)

the stable channel was updated to 39.0.2171.99. Here are the md5sums: md5sums=('037cef95bea64ffa87e926af59ae117b' 'b7e752f549b215ac77f284b6486794b6')

derrickcope commented on 2015-01-07 04:41 (UTC)

since the last upgrade it keeps freezing up on me. Mainly when I turn on/off the vpn. never happened before.

feighur commented on 2014-12-23 01:39 (UTC)

Even with ttf-dejavu and ttf-liberation installed, I had Chinese characters show up as boxes. Tested on I had to install ttf-droid to get them to display properly. Perhaps another optional dependency?

rafaelff commented on 2014-12-10 15:18 (UTC)

@grayich: This is related to your configuration in /etc/yaourtrc (or maybe in /etc/makepkg.conf), where you set to store built packages in /var/cache/pacman/pkg/ Set another folder where you can store without needing root permissions, or always provide password. Anyway, this issue is not related to this package, so further issues, ask for help in Arch Linux forums.

Det commented on 2014-12-10 11:50 (UTC)

Nothing at all to do with this package.

Det commented on 2014-12-09 19:34 (UTC)

The servers still offer .71 for Linux:

Det commented on 2014-11-28 17:45 (UTC)

Yes, it works on a fresh install. About the wrong source, you're probably using a poor AUR tool, or the system really is 64-bit: $ uname -m $ grep CARCH /etc/makepkg.conf

Av4t4r commented on 2014-11-28 17:40 (UTC)

I just did a fresh install with no permission issues. However, I did it on a 32bit system and I had to edit the PKGBUILD because it would otherwise only fetch the 64 bit package. I edited _arch=i386 and updated the MD5 to md5sums=('cc7c7cfcbd9982307b25989814739486' 'b7e752f549b215ac77f284b6486794b6') Thanks for the package tho, hopefully this will get fixed soon.

Det commented on 2014-11-27 13:33 (UTC)

No, that's not the issue at all, aren't you getting the permission message after the upgrade? E.g.: [...] (1/1) upgrading google-chrome [#####################] 100% warning: directory permissions differ on /usr/share/doc/google-chrome-stable/ filesystem: 700 package: 755 It means the existing permissions can't be changed without a manual intervention or a reinstall (uninstall/install).

nullptr_t commented on 2014-11-27 13:01 (UTC)

I know, you are fixing the permissions of the doc directory in PKGBUILD with msg2 "Fixing permissions of documentation folder" chmod 755 "$pkgdir"/usr/share/doc/google-chrome-$_channel/ but as of 39.0.2171.71-1 the problem still remains the same: # ll -d /usr/share/doc/google-chrome-stable/ drwx------ 2 root root 4.0K Nov 27 13:46 /usr/share/doc/google-chrome-stable/ Don't know, what's the problem with "google-chrome-$_channel/", but maybe "$_channel" or ${_channel} (like in the sources-line) would help to clarify it as substring

jazztickets commented on 2014-11-26 03:19 (UTC)

google-chrome-stable_39.0.2171.65_amd64.deb ... FAILED ERROR: One or more files did not pass the validity check!

rafaelff commented on 2014-11-13 18:10 (UTC)

I'm getting http error 404 on downloading the sources tarball.

Det commented on 2014-11-03 18:43 (UTC)

Apparently a pretty old bug. Couldn't reproduce with Beta or Dev channels:

CarstenG commented on 2014-11-03 18:28 (UTC)

Please update the permissions of the doc directory, because unprivileged users can't access the content. # ll -d /usr/share/doc/google-chrome-stable drwx------ 2 root root 25 31. Okt 17:42 /usr/share/doc/google-chrome-stable Thanks, Carsten

cyrilccros commented on 2014-10-27 21:39 (UTC)

Hi, I just ran the usual pacman -Syu today and all my troubles went away. I still had the bug this morning, though. Anyway, thanks for your help.

cyrilccros commented on 2014-10-27 09:02 (UTC)

Sorry for the delay in my answer. I have spent some time googling my issues, but google's error messages are of little use. I see several different causes for the same message... See below for a full error example: [cyril@synapse ~]$ google-chrome-stable ATTENTION: default value of option force_s3tc_enable overridden by environment. [17667:17735:1027/] PostClientToServerMessage() failed during GetUpdates [17667:17701:1027/] RawChannel read error (connection broken) Bus error (core dumped) What I do remember is that my computer crashed very shortly before my troubles began. I use Openbox + Slim as login manager, and slim failed to launch openbox repeatedly. I just used the console to log in and correct slim.conf, which solved my issues.

drak3 commented on 2014-10-25 18:41 (UTC)

I, too, am having issues with Chrome 38 crashing frequently. It doesn't do it for every page, but it does it frequently, even after disabling all my extensions, re-installing, and making a new profile. after a few minutes, it will also lock up the system, requiring a hard reboot. also, as with cyrilccros, firefox seems to be fine. one thing I've noticed that will definitely trigger the tab crash is playing video in multi-monitor fullscreen.

Det commented on 2014-10-22 22:59 (UTC)

Did you try first?

cyrilccros commented on 2014-10-22 22:56 (UTC)

Hi, google-chrome crashes whenever I load a page. This happens with chromium too, but Firefox is fine. I get an error in] PostClientToServerMessage() failed during GetUpdates Bus error (core dumped) Chrome also crashes shortly after start up, and I often get "Aw snap" error windows. I tried launching it with no plugins or gpu support. I use Bumblebee (Intel and Nvidia laptop). Thanks for your help, I will post a bug report if anyone confirms. Version: google-chrome=38.0.2125.104 (l test)

Det commented on 2014-10-20 13:31 (UTC)

It's now an optional dependency.

catalin.hritcu commented on 2014-10-18 13:27 (UTC)

Chrome doesn't display many PDFs correctly without ttf-liberation ( Shouldn't that be a package dependency?

Det commented on 2014-10-16 11:16 (UTC)

Please Google first: (E.g. the very first results).

zlopi commented on 2014-10-16 08:40 (UTC)

NVIDIA: could not open the device file /dev/nvidia0 (Operation not permitted). NVIDIA: could not open the device file /dev/nvidia0 (Operation not permitted). how to fix it?

abhixec commented on 2014-10-15 04:09 (UTC)

Updated PKGBUILD :

yodalee commented on 2014-10-15 02:43 (UTC)

The hash of google-chrome.deb should be a2b61a623da973f12e1cb18674b9fb93 now

lviggiani commented on 2014-10-08 07:32 (UTC)

Every time I get an update of the package, Google complains that cannot load my profile. Even by re loggin in does not solve the isssue. I have to delete ~/.config/google-chrome folder and re do the access to all my profiles. This is happening since the last three or four updates...

Det commented on 2014-10-01 05:32 (UTC)

Same as with all the other OS's:

Barry commented on 2014-10-01 03:07 (UTC)

What's the difference between chrome and chromium on Arch?

Det commented on 2014-09-25 11:42 (UTC)

@corvinusz: "The stable channel has been updated to 37.0.2062.124 for _Windows_ and _Mac_." -

capoeira commented on 2014-09-21 13:02 (UTC)

Oh I see. now I just need to convince my bank to stop using it lol

Det commented on 2014-09-21 12:43 (UTC)

No it's not blocked, the whole support for the plugin API is gone.

capoeira commented on 2014-09-21 11:29 (UTC)

i understand that Java is blocked...the thing is, it's not even in chrome://plugins/

Det commented on 2014-09-20 20:09 (UTC)

It (NPAPI support) has been gone from Linux builds since v35:

capoeira commented on 2014-09-20 20:00 (UTC)

for the last 2 builds java doesn't apear anymore in about:plugins.......anyone?

cleanrock commented on 2014-09-12 12:43 (UTC)

PKGBUILD is broken for i686: PKGBUILD: line 28: md5sums[0]: cannot assign list to array member Just remove the (), line should be like this: [[ "$CARCH" = 'i686' ]] && md5sums[0]='fc4fd42aa00222a7df9ad89417f8d219'

x7.0 commented on 2014-09-12 01:46 (UTC)

"an error in package()" may be caused by SSL certificate error in line 44-45 in PKGBUILD, commenting out would help. I wonder if it is possible to put the license out of `package()`, because in China we cannot connect to Google directly without any proxy.

Det commented on 2014-09-11 14:40 (UTC)

I really am unable to help with that.

passchaos commented on 2014-09-11 08:17 (UTC)

Help!!When I build chrome, I get "an error in package()"

phoenixlzx commented on 2014-09-11 02:48 (UTC)

Hi, Why the PKGBUILD says its `37.0.2062.120-1` but title shows `37.0.2062.94-1`? BTW version `37.0.2062.120-1` md5sum: x86_64: ad13b2cd986933a57cf348dc0efc50cf i686: fc4fd42aa00222a7df9ad89417f8d219

Det commented on 2014-09-06 22:07 (UTC)

No problem bringing it up. I didn't even know it was a Chromium bug, but usually these aren't Chrome-specific. Otherwise these bugs would be closed anyway, because that's code we don't get to see. p.s. I appreciate you liking my writing. I like it as well.

commented on 2014-09-06 03:34 (UTC)

@Det I really like the way you reply to people on the packages you maintain (no sarcasm, I really do like it). Anyway, by "anyone here" I meant someone running Arch and using this package (or maybe chromium). I got this bug initially when M37 was on Dev and it still wasn't fixed, so I just wanted to point to the bug report in case someone got Chrome hanging for "no apparent reason" or for the somewhat easily, and almost certainly annoyingly, visible tooltip. The stack trace on comment #10 is probably enough to fix, but if anyone else gets any useful information it might be helpful to comment there. Or star it if you just want to get updates.

Det commented on 2014-09-05 23:32 (UTC)

Yeah. It was confirmed on Fedora also in your report. Also using Gnome Shell.

commented on 2014-09-04 17:11 (UTC)

Anyone here also affected by this issue: ?

keiseele commented on 2014-08-31 19:08 (UTC)

fix the problem of chrom UI after install gnome-themes-standard, his package also install ttf-dejavu, after install chrome work fine and all menu and UI fonts are fine

keiseele commented on 2014-08-31 19:00 (UTC)

my chrome and chromium have bad UI fonts, I update my system with pacman -Syu and also update chrome to last stable version. here a screenshot of my chrome, notice the tab name, url and menu are completely unreadable, some help please. system kde and video driver is intell

Det commented on 2014-08-13 12:36 (UTC)

Already updated.

niklas.heer commented on 2014-08-13 12:35 (UTC)

New md5sum: "b00bd4f59950583da013fc79e06db340"

Det commented on 2014-07-28 12:03 (UTC)

Removed already from -beta and -dev.

hadrons123 commented on 2014-07-28 05:16 (UTC)

libgcrypt dependency has been removed in chrome upstream. can you remove it from dependency list at the next release bump?

Det commented on 2014-07-21 20:51 (UTC)

Fabulous, but next time preferably Pastebin.

weirddan455 commented on 2014-07-20 22:02 (UTC)

Seems I've gotten pretty good at solving these Chrome UI bugs. I'll copy my fix from the Chromium bug tracker to help my fellow Arch users out. OK I've found the root of this bug and it's not really Chrome's fault. There's a dirty hack in the Adwaita theme to work around a bug in an old version of LibreOffice. To fix, look in /usr/share/themes/Adwaita/gtk-2.0/gtkrc and look for this section of code: style "menu_item" { xthickness = 2 ythickness = 4 # HACK: Gtk doesn't actually read this value # while rendering the menu items, but Libreoffice # does; setting this value equal to the one in # fg[PRELIGHT] ensures a code path in the LO theming code # that falls back to a dark text color for menu item text # highlight. The price to pay is black text on menus as well, # but at least it's readable. # See bg[SELECTED] = @selected_fg_color bg[PRELIGHT] = @selected_bg_color fg[PRELIGHT] = @selected_fg_color text[PRELIGHT] = @selected_fg_color Change bg[SELECTED] = @selected_fg_color to @selected_bg_color and the bug should be fixed for Chrome. Not sure if this breaks recent versions of LibreOffice or not. The reason this only affects Chrome is that every other GTK application (with the exception of LibreOffice apparently) uses the bg[PRELIGHT] field while Chrome uses the bg[SELECTED] field. The reason this broke with version 36 is there was also a section in that same file that reverted this behavior specifically for Chrome. style "chrome_menu_item" { bg[SELECTED] = @selected_bg_color } And further down where this is applied: # Chrome/Chromium class "ChromeGtkFrame" style "chrome-gtk-frame" widget_class "*Chrom*Button*" style "button" widget_class "*<GtkCustomMenu>*<GtkCustomMenuItem>*" style "chrome_menu_item" # We use this weird selector to target an offscreen entry as created # by Chrome/Chromium to derive the style for its toolbar widget_class "<GtkEntry>" style "standalone_entry" I'm guessing Google changed the name of its frame from ChromeGtkFrame (or maybe they changed the name of their menu.) In any case, some change to Chrome broke this reverse-dirty-hack and caused the bug.

commented on 2014-07-18 17:46 (UTC)

This bug now also affects the Stable channel:

Det commented on 2014-07-10 19:04 (UTC)

No I mean as in it's not necessary to put in "Don't believe me?" in the midst of all that, when you clearly know what you're talking about.

weirddan455 commented on 2014-07-10 19:02 (UTC)

...What don't you believe? Did setting XCURSOR_THEME not fix Chrome's cursor for you?

Det commented on 2014-07-10 18:54 (UTC)

I _don't_ believe you.

weirddan455 commented on 2014-07-10 16:01 (UTC)

On the mouse cursor issue, the reason no one is fixing it is because it's not a Chrome bug. You should be filing a bug report for your DE for not correctly setting the XCURSOR_THEME environment variable (and optionally XCURSOR_SIZE if you need a custom size) That's the standard way for every X application to know which cursor to use. On KDE, this variable is set in the /usr/bin/startkde script. Don't believe me? Try launching any X application from the console, setting that variable to a different font. For example "XCURSOR_THEME=Oxygen_Blue firefox" Firefox used to use Oxygen_White like the rest of the system but now it uses Oxygen_Blue. Chrome works the same way. The old version may have grabbed your cursor theme setting from the GTK theme but since the new version no longer uses GTK it's not going to look there. On the whole screen going blank, I would try disabling GPU acceleration. In Settings > Advanced, uncheck "Use hardware acceleration when available." You can also disable Flash's GPU acceleration by right clicking a Flash widget and clicking settings. You could also just disable Flash in about:plugins to rule it out entirely (may want to try this first.)

catalin.hritcu commented on 2014-06-26 08:19 (UTC)

Managed to find something in dmesg that seems correlated with the failures: My video card is a Intel HD Graphics 4600, on a Lenovo T440p. There is also an NVidia card inside but it's disabled.

catalin.hritcu commented on 2014-06-26 08:01 (UTC)

I'm experiencing a new problem with the latest version: every now and then the whole screen starts going blank for a while. Not just the Chrome window but the whole screen. And it doesn't even matter whether the Chrome window is active or just somewhere in the background. All I could find to stop the problem is to restart Chrome, but eventually it comes back. I guess it's related to the new Aura thing, but I didn't manage to get any output in the console that would give me more information to be able to file a good bug report. I'm using MATE.

Det commented on 2014-06-23 19:47 (UTC)


mityukov commented on 2014-06-23 19:11 (UTC)

Could you add $CHROMIUM_USER_FLAGS support? E.g., by modifying the exec line in the end of google-chrome-stable: ... exec -a "$0" "$HERE/chrome" $CHROMIUM_USER_FLAGS "$@" ... (I'm not sure this is the proper way yo do so) It would be very helpful if it goes with the package.

Det commented on 2014-06-22 18:03 (UTC)

Lol: "I am not planning on working on this soon. Putting back into the queue."

commented on 2014-06-22 18:00 (UTC)

@catalin.hritcu I reported the mouse cursor problem some time ago: Looks like nobody is currently working to fix it.

catalin.hritcu commented on 2014-06-22 15:43 (UTC)

Seems that the display artifacts are gone (for me at least) starting in 35.0.1916.153. The mouse cursor looks strange though (not respecting the desktop theme I've set in MATE). Is there any way to tell Chrome to use the standard mouse cursor icons instead of its own ones?

whynothugo commented on 2014-06-18 06:46 (UTC)

@lockheed: chrome (and chormium) have recently moved to their own toolkit (aura), which is still quite unpolished. That MAY be realted to your issue.

matkam commented on 2014-06-11 03:23 (UTC)

@Det: Thanks. Turned out I had a couple old desktop files in my home dir.

KevinCodux commented on 2014-06-10 19:13 (UTC)

New hashes: i386: 2e51d1d06b6fdcb95b25042be0b44300 amd64: 0cf3cbd05f4a8c90e88b597b0c4028a0

KevinCodux commented on 2014-06-10 19:04 (UTC)

Current MD5/checksum is invalid

lockheed commented on 2014-06-10 15:05 (UTC)

Since few days back, I no longer have ability to type non-english letters in Chrome (and only in Chrome). I mean letters like ą, ę, ć, ź etc. What could be causing that?

Det commented on 2014-06-10 07:03 (UTC)

└┌(%:~)┌- grep Exec /usr/share/applications/google-chrome.desktop Exec=/usr/bin/google-chrome-stable %U Exec=/usr/bin/google-chrome-stable Exec=/usr/bin/google-chrome-stable --incognito

huorpalantir commented on 2014-06-04 07:44 (UTC)

Fonts and icons seem to be changed (smaller/larger/reverted to default) in Chrome 35 compared to Chrome 34. Folks who value their font size or icons shouldn't update.

Det commented on 2014-06-04 02:04 (UTC)

Well, I did it for you: Anyway, I see nothing wrong with your system, so it's either got something to do with your specific configuration or Antergos in general. Either way, I just can't help you.

Det commented on 2014-06-04 01:49 (UTC)

Pastebin, please :D.

Det commented on 2014-06-04 01:35 (UTC)

I'm glad you're being helpful, but that's pretty obvious :).

commented on 2014-06-04 01:34 (UTC)

Just one idea about the udev problem: Maybe usb_modeswitch or other package had the symlink needed and either dropped it or was uninstalled? I am not sure why it is not included in the udev(maybe systemd) package, it would solve a lot of these issues.

Det commented on 2014-06-04 01:32 (UTC)

Since this is obviously important to you, would you mind running: $ ls -l /opt/google/chrome/ /usr/lib/libudev*

Det commented on 2014-06-04 00:50 (UTC)

Then that's a problem with Antergos. Not with this.

Det commented on 2014-06-04 00:30 (UTC)

1) I don't get that error. 2) You don't have to repost to include "package flagged", because I'm already getting notified from all of these.

starnostar commented on 2014-06-03 00:20 (UTC)

I tried this after chromium (stable) v35 started acting funny... I keep having the same issue where I cannot drag and drop images to download them. I'm really kind of astounded that there is no "save image" right click action (only "save image as..") and no extensions that emulate this functionality.

Det commented on 2014-06-02 10:55 (UTC)

@nicoazrak, done.

Det commented on 2014-05-31 00:54 (UTC)

At least on Gnome Shell the only thing I've noticed is the graphical corruption and sluggishness when resizing the window. With Nvidia 337.25 the sluggishness is also gone:

victorz commented on 2014-05-30 18:57 (UTC)

I find it strange that so many are having problems with v35. It works fine on other distros (eg Debian). Either way, just wanted to remind everyone to also report bugs to the Chromium bug tracker. :-) Thanks! (

heaven commented on 2014-05-30 17:29 (UTC)

I was using v35 all this time since the update and can conclude it is total cr*p, if devs are going to continue this way I'd definitely need another browser.

Det commented on 2014-05-29 22:10 (UTC)

Update: v34 is a flag on top of the PKGBUILD for now.

krix commented on 2014-05-29 19:07 (UTC)

@Det Thank You very Much for 34 PKGBUILD! :) pipelight works great now.

Det commented on 2014-05-27 09:59 (UTC)


max commented on 2014-05-27 07:48 (UTC)

With the latest version 35.0.1916.114 i can't use external plugins (java, etc) chrome://plugins show only the defaults, others plugins are not present (even if they are correctly installed) Before update they worked.

nicoazrak commented on 2014-05-27 01:26 (UTC)

If somebody whats to have chrome pinned to docky add StartupWMClass=Google-chrome-stable to the .destktop file. Is it posible to add it to the .deb file? More info:

Det commented on 2014-05-23 19:48 (UTC)

And, as also mentioned, at least one possibility is your passwords.

catalin.hritcu commented on 2014-05-23 19:08 (UTC)

Downgrading to Version 34.0.1847.137 causes this warning: Your profile can not be used because it is from a newer version of Google Chrome. Some features may be unavailable. Please specify a different profile directory or use a newer version of Chrome.

Det commented on 2014-05-23 18:59 (UTC)

Bumping v34 for anyone else not reading the comments:

catalin.hritcu commented on 2014-05-23 18:55 (UTC)

I get strange graphics artifacts starting with Version 35.0.1916.114.

reed1 commented on 2014-05-22 01:50 (UTC)

thx Det! v35 was very awful, SmoothScroll also did not work.

Det commented on 2014-05-21 18:52 (UTC)

Those with LastPass don't have these kinds of issues :).

heaven commented on 2014-05-21 18:50 (UTC)

After rollback it isn't able to work properly with saved passwords (doesn't fill inputs with saved values). Perhaps they were converted to another format.

heaven commented on 2014-05-21 18:30 (UTC)

@Det, yup, didn't notice you changed the source url in pkgbuild. That worked for me, thank you. Also noticed 35 version doesn't work properly with gtk-oxygen (using in KDE).

Det commented on 2014-05-21 15:22 (UTC)

What we've been talking about the whole time?

codenamestif commented on 2014-05-21 15:21 (UTC)

What's happend with new google chrome view? There are some problems with fonts and with rendering ui at all

mssun commented on 2014-05-21 12:17 (UTC)

When I open a chrome instance, the page is in full screen mode, I cannot exit. What I can do is just hit ALT+F4 to close the window and open a new one. I am using cinnamon.

Det commented on 2014-05-21 11:14 (UTC)

I just gave you the link for the PKGBUILD?

heaven commented on 2014-05-21 11:13 (UTC)

@Det, ohh, thank you, I see now. Is there any way to get deb with 34.0.1847.137?

Det commented on 2014-05-21 05:59 (UTC)

The previous version wouldn't help you, because there's only one link for each channel, but: We also do snapshots, so you don't have to try to remember what you've done in the past, or ask other people for copies:,

heaven commented on 2014-05-21 05:48 (UTC)

@Det, hi, do you have the previous PKGBUILD version for the 34 version? 35 does not look like a stable release. I feel some performance issues, tabs switching is slow, some rendering issues. Chrome team frustrates me more and more with each new release.

Det commented on 2014-05-21 05:04 (UTC)

@lucaskatayama, anyone? _Everyone_:

cellsheet commented on 2014-05-21 00:03 (UTC)

chrome version 35 has minimization issues where the buffer isn't refreshing and leaves chrome visible on desktop. It also re-draws anything animating (youtube, flash, gif's, etc) over the whole desktop space. Also tool tip text is white on yellow when using black mate theme.

lucaskatayama commented on 2014-05-20 23:49 (UTC)

It's not finding java plugin. Anyone with the same problem?

Det commented on 2014-05-20 19:51 (UTC)

└┌(%:~)┌- google-chrome-stable /usr/bin/google-chrome-stable: error while loading shared libraries: cannot open shared object file: No such file or directory I disagree with you.

Det commented on 2014-05-20 19:45 (UTC)

What if I refuse?

vorot93 commented on 2014-05-20 19:43 (UTC)

Please remove gtk2 dependency: Chrome uses Aura by default now.

mssun commented on 2014-05-13 06:10 (UTC)

When opening a PDF file, all pages crash. After refreshing the page, I can view the PDF in browser without any problem. Any one have the same problem?

ddracopo commented on 2014-05-11 16:11 (UTC)

In most of the pages, minimising chrome still leaves parts of its pages rendered on the screen. One needs to refresh the window manager to get rid of that, but the problem occurs again when minimising chrome. I use icewm as the window manager. Firefox does not have this problem. Anyone else who has the same problem?

sflor commented on 2014-05-11 14:22 (UTC)

@mityukov Maybe your graphics driver at fault. Have you tried disabling hardware acceleration? There are also ways to go back to flash if that does not help. Just google it.

mityukov commented on 2014-05-06 12:05 (UTC)

When trying to watch video with HTML5 player (it's by default for YouTube videos and I don't know how to turn it to Flash), the page gets "Aw Snap!" at the very first second.. Are there any packages (codecs/etc.) I should check/install on my system? P.S.: I've started getting this problem no longer than a month ago.

Det commented on 2014-04-27 15:49 (UTC)

Then it's a _B_rackets bug and for me to hardcode a symlink would make the Chrome packages conflict with each other again after Google just separated them.

ansidev commented on 2014-04-27 12:43 (UTC)

If you use brackets on your system, you will see an error when trying to use Live Preview feature. You must run command: sudo ln -s '/usr/bin/google-chrome-stable' '/usr/bin/google-chrome' to fix this error. It's also similar to other Google Chrome 's channel. @Det: I hope you to add this command to the package.

mike.cloaked commented on 2014-04-25 08:52 (UTC)

Det: we are all human!

Det commented on 2014-04-25 08:45 (UTC)

Broke it pretty quickly too.

mike.cloaked commented on 2014-04-25 08:44 (UTC)

Thank you for fixing the script so quickly.

Det commented on 2014-04-25 08:02 (UTC)

Calm down already. I have all these things in my own scripts directory and that's where 'msg2' was picked up instead, which made me think it wasn't necessary to define it in the .install. Turns out it is, so I've reverted it to 'msg_blue', which is more consistent with 'note'. All your cache commands were working okay. The printf() function changes wouldn't affect them because they are basically just separate echoes.

mike.cloaked commented on 2014-04-25 07:56 (UTC)

neither msg2 nor msg_blue work so I guess the available commands in the makepkg system may have changed?

mike.cloaked commented on 2014-04-25 07:50 (UTC)

previously in the file google-chrome.install where the two lines had msg2 they used to have msg_blue.

mike.cloaked commented on 2014-04-25 07:47 (UTC)

hadrons123: There was until this version instead of the two lines you quote: ==> Updating desktop MIME database... ==> Updating icon cache.. So I guess it is in the file google-chrome.install where the lines: _update() { msg2 "Updating desktop MIME database..." update-desktop-database -q msg2 "Updating icon cache.." gtk-update-icon-cache -q -t -f /usr/share/icons/hicolor are no longer accepting the msg2 command, but presumably the update-desktop-database and gtk-update-icon-cache are still working OK.

hadrons123 commented on 2014-04-25 04:30 (UTC)

/tmp/alpm_hA23f7/.INSTALL: line 12: msg2: command not found /tmp/alpm_hA23f7/.INSTALL: line 14: msg2: command not found ==> NOTE: The binary is called 'google-chrome-stable'. [root@archbox ~]# Any idea why I am getting this error today?

Det commented on 2014-04-21 19:16 (UTC)

@dmoebius, sounded like you just did.

Det commented on 2014-04-21 17:11 (UTC)

Taken care of.

dmoebius commented on 2014-04-21 17:04 (UTC)

Yes, just now.

Det commented on 2014-04-21 16:18 (UTC)

Did you ever redownload it?

dmoebius commented on 2014-04-21 14:00 (UTC)

Since half a year or so, the md5sum for eula.txt is 6d57da7476a4b1b7a81821d9c036425c, not ba19ea498f294975d320ff0b26a6cd63. When will this be fixed?

Det commented on 2014-03-28 15:39 (UTC)

The depends=() would be unnecessarily long, if I were to include all requested dependencies' dependencies (which are already satisfied). It would be something like: depends=('alsa-lib' 'atk' 'attr' 'avahi' 'bzip2' 'cairo' 'dbus-glib' 'desktop-file-utils' 'expat' 'flac' 'fontconfig' 'freetype2' 'gconf' 'gdk-pixbuf2' 'glib2' 'graphite' 'gtk2' 'harfbuzz' 'harfbuzz-icu' 'hicolor-icon-theme' 'icu' 'keyutils' 'krb5' 'libcap' 'libcups' 'libdbus' 'libffi' 'libgcrypt15' 'libgl' 'libgpg-error' 'libpng' 'libx11' 'libxau' 'libxcb' 'libxcomposite' 'libxcursor' 'libxdamage' 'libxdmcp' 'libxext' 'libxfixes' 'libxi' 'libxinerama' 'libxrandr' 'libxrender' 'libxss' 'libxtst' 'nspr' 'nss' 'openssl' 'opus' 'pango' 'pcre' 'pixman' 'snappy' 'speech-dispatcher' 'ttf-font' 'xdg-utils' 'zlib') opposed to (which installs all of these): depends=('alsa-lib' 'desktop-file-utils' 'flac' 'gconf' 'gtk2' 'harfbuzz' 'harfbuzz-icu' 'hicolor-icon-theme' 'icu' 'libgcrypt15' 'libpng' 'libxss' 'libxtst' 'nss' 'opus' 'snappy' 'speech-dispatcher' 'ttf-font' 'xdg-utils')

brfransen commented on 2014-03-26 15:09 (UTC)

libcups should be a depends.

Det commented on 2014-03-04 13:51 (UTC)

Why? Are you from the future?

ushi commented on 2014-03-04 13:50 (UTC)

Integrity check failed for google-chrome-stable_33.0.1750.146_amd64.deb

Det commented on 2014-03-03 22:23 (UTC)

Well, of course it fucking is, as soon I get my LAN cut off they upload a new build.

yetr2 commented on 2014-03-03 22:19 (UTC)

md5sum for i686 package appears to be incorrect # md5sum ./google-chrome-stable_33.0.1750.117_i386.deb 052ab0591558644afd459507a968e462 ./google-chrome-stable_33.0.1750.117_i386.deb

heliumhgy commented on 2014-03-02 03:42 (UTC)

welp, sometime I see this ERROR:] sendmsg: Broken pipe

heliumhgy commented on 2014-03-02 03:38 (UTC)

I have CPU usage issue. Chrome renders way slow and eats up too much CPU. for test: (or even much much slower and CPU intensive) add it to bookmarks then open multiple tabs of it, check the task manager and I found over 100% usage by the "Browser" process, and all tabs are blank, it takes intolerable long time to load the page and makes my system slow. (preceding issue also happens in Chromium) If I use Firefox to load above URL in multiple tabs simultaneously, it loads almost instant up to my network capacity. If I run google-chrome-stable from command line, I saw no error or warning in prompt. And I tested the newest version Chrome in an openSUSE 12.3 and it suffers the same problem but with much low CPU usage. Is it a Chrome bug or Linux issue or arch's problem? Can anyone helps me with it?

Det commented on 2014-02-24 05:16 (UTC)

Removed "PKGEXT='.pkg.tar'" due to a request of having your own way through makepkg.conf.

Det commented on 2014-02-23 17:42 (UTC)

On the KDE issue, I don't know, if it's because of '/usr/share/nvidia/nvidia-application-profiles-334.16-rc' or some other update in general, but so far Chrome hasn't frozen on me and seems extremely snappy. I may just be able to return to my dearest KDE.

waders commented on 2014-02-21 00:18 (UTC)

Fix it then pastebin it, help the Maintainer. Or install it yourself.

Det commented on 2014-02-20 23:22 (UTC)

Yeah, I know they have concerns, but it's not the end of the world and something everybody has to comment on. You flag it once, and it won't be updated any faster no matter how many times you wonder why it doesn't match.

Archwyrm commented on 2014-02-20 23:14 (UTC)

Det: I ran into the md5sum issue with the older google-chrome-stable_32.0.1700.107_amd64.deb so perhaps others had some concern along those lines. Personally, since the file name was the same and only the md5sum differed, I decided to play things safe and wait for an update. Only had to wait a day. Thank you for your efforts! :) A little patience goes a long way..

mike.cloaked commented on 2014-02-20 21:57 (UTC)

Only seconds after I looked it was already fixed... thank you Det (I presume).

mike.cloaked commented on 2014-02-20 21:53 (UTC)

It is simple to edit the PKGBUILD file with the new md5sums - but also you need to change the build number to end in 117 - the current file ends in 115. Other than that it is fine.

Det commented on 2014-02-20 21:47 (UTC)

I can't believe you guys raise such a fuzz about an outdated PKGBUILD.

rikless commented on 2014-02-20 19:47 (UTC)

concerns by FadeMind issue too.

FadeMind commented on 2014-02-20 19:10 (UTC)

Today Google released second update of Google Chrome. Current version is 33.0.1750.117 Correct MD5SUMS: $ md5sum google-chrome-stable_33.0.1750.117_i386.deb 8159f0484cf3fab3e768d1665aa58871 google-chrome-stable_33.0.1750.117_i386.deb $ md5sum google-chrome-stable_33.0.1750.117_amd64.deb a5c986f948ca72db015236fb9b6bb986 google-chrome-stable_33.0.1750.117_amd64.deb PKGBUILD MD5SUM CORRECT: md5sums=('8159f0484cf3fab3e768d1665aa58871' 'ba19ea498f294975d320ff0b26a6cd63') [ "$CARCH" = 'x86_64' ] && md5sums[0]='a5c986f948ca72db015236fb9b6bb986' AUR now have outdated data in PKGBUILD file.

A.J commented on 2014-02-20 18:50 (UTC)

the stable channel is at 33.0.1750.117 ??!

keepitsimpleengr commented on 2014-02-20 17:58 (UTC)

anybody know the md5sums for google-chrome 32bit ?

silent commented on 2014-02-20 17:29 (UTC)

md5sums for google-chrome amd64 Version 33.0.1750.117 should be 'a5c986f948ca72db015236fb9b6bb986'

Det commented on 2014-02-14 19:05 (UTC)

Removing 'libcanberra-pulse' was reported a long time ago to fix the issue: I don't know, if PulseAudio needs to be removed too.

reed1 commented on 2014-02-14 17:45 (UTC)

Thx @Det, I left KDE for Openbox then everything went well. tip for OB: use compton not xcompmgr

Det commented on 2014-02-08 16:33 (UTC)

I used to get it with KDE all the time and got incredibly frustrated with both pretty much no one else noticing and updates helping shit. Hasn't happened with GNOME, though, I still prefer KDE.

reed1 commented on 2014-02-08 16:08 (UTC)

Anyone got random crash / segment fault? I got it far too often for a browser :( disabling extension / deleting profile folder won't help...

Det commented on 2014-02-03 20:26 (UTC)

Works fine here.

Det commented on 2014-01-31 14:14 (UTC)

Remove /usr/lib/ (or /opt/google/chrome/ first. Chrome doesn't use the 1.6 you get from the repos (

leduck commented on 2014-01-31 08:44 (UTC)

I'm not having problems with Chrome and am using libgcrypt 1.6.0-2, why should I install the old library?? I use the x64 version of Arch ...

jayflo commented on 2014-01-23 00:00 (UTC)

Chrome sometimes hangs when I try to copy/paste from a web page. Usually pages that have forms or that allow editing documents (like on github editing a readme). Is this happening for anyone else?

netsurfer commented on 2014-01-16 14:43 (UTC)

i have installed the libgcrypt15 and snappy (in my case) whit pacman and then installed google-chrome tarball; now all seems ok, the patch for what purpose is created?thanks for your effort

Det commented on 2014-01-16 10:23 (UTC)

We're awesome, aren't we?

justinkb commented on 2014-01-16 10:01 (UTC)

Always thought AUR was the perfect example of the potential of what relatively unorganized community effort can achieve ;-)

Det commented on 2014-01-16 07:00 (UTC)

I love you guys always handing out these patches to change a couple of lines :).

justinkb commented on 2014-01-15 23:36 (UTC) Simple patch to use the libgcrypt15 as suggested below

r2b2 commented on 2014-01-15 20:16 (UTC)

Now that there is a libgcrypt15 package in AUR for providing the old libgcrypt libraries, would it make sense to provide a dependency back to that? (Or at least reinstate the commented if statement which checks for libgcrypt and check for libgcrypt15 before copying the included libgcrypt libraries?)

l3u commented on 2014-01-15 08:19 (UTC)

Thanks for the update Det, and thanks for the work you're doing on it. I see a some people dissatisfied with the free service you're providing to the community on your own time. Let me say I am delighted to have an up to date system thanks to dedicated people like you. Deps change. Glitches happen. This is Arch. We have to live with this. If edges are too rough, there are other OSes out there. At least Arch let's you get the original PKGBUILD & friends and fix the problem by yourself, which is what I do when things go wrong, instead of pointing fingers. Life is hard for package maintainers. Behaving is the minimum we owe them.

Det commented on 2014-01-15 00:22 (UTC)

Please say this again; I loved this: > Comment by Scimmia 2014-01-14 17:44 > You failed. Bear with me here, darling. I didn't think this through. This wasn't the sort of running issue as you get with source packages where you first require a rebuild for [testing] users and then, a few weeks later, for everybody else. Chrome is a binary package. It _can't_ be rebuilt. I'm putting the actual library in, not recompiling for it. This could've _indeed_ (or rather should've) been done those 3 weeks ago, so now that non-testing users are also getting the update they'd already have a working Chrome. Well, apologies. Sometimes I think of some common solution straight away that doesn't actually apply to the current case at all. I'll try to be better?

lrm commented on 2014-01-14 20:51 (UTC)

The deb fails validation on 64-bit build. New checksum is: 5eda4b45f5482e526252a819ed5e78a3

w1ntermute commented on 2014-01-14 20:13 (UTC)

I agree with the other commenters, you really should have incremented pkgrel. That's the reason why pkgrel exists, like Scimmia said. OK, you made a mistake at first, but there's no reason to keep digging a deeper hole for yourself - just admit that you made a mistake and don't let it happen again.

nDray commented on 2014-01-14 20:07 (UTC)

And if you must know, pkgrel should have been incremented

nDray commented on 2014-01-14 20:07 (UTC)

New version is out, just put people out of their misery with the update.

Scimmia commented on 2014-01-14 17:44 (UTC)

The real problem is that you didn't bump the pkgrel when you fixed the package in the first place. If the final binary package changes, you're supposed to bump the pkgrel. You failed.

big_bum commented on 2014-01-14 17:06 (UTC)

It will really be easier for all of us to simply increase the pgkrel. You are not rebuilding anything, you just have to edit one file...

Det commented on 2014-01-14 16:58 (UTC)

Unfortunately there's no such pkgrel that only affects those who have a "broken" system. When something that breaks things moves from [testing] to [core]/[extra] I'm against every single step of raising the pkgrel on any package that's not small as hell. It would have to be something seriously major and AUR rarely counts as such. That's life, that's Arch. If you don't have those libraries yet, then rebuild it or manually put them there. Those who do don't care.

Scimmia commented on 2014-01-14 16:39 (UTC)

nagging? It's required to rebuild it with the new libgcrypt, not nagging to help keep a working system.

Det commented on 2014-01-14 16:06 (UTC)

Not really. I did that change a long time ago and don't care about nagging anybody to rebuild it.

ushi commented on 2014-01-14 12:44 (UTC)

you could increase the pkgrel number. this is exactly the use case.

Det commented on 2014-01-13 21:37 (UTC)

Roughly, no, because for one, I would not bump the _version_ number to do that (which is '31.0.1650.63'), it would be the _release_ number (that is, '1') and two, the new libgcrypt 1.6.0-1 is the reason Google Chrome/Chromium/SRWare Iron do not work without the previous libraries being put in place (which is what I'm currently doing for you). Chrome is a binary package. I can't just go forward with my ambitious plans to rebuild it for a newer library that has been pulled in on another distribution, totally different fromt the ones they support. To sum up: - Chrome is built for libgcrypt 1.5.x - We use libgcrypt 1.6.x - They do NOT work together - Only Google can fix this - We (I) use the 1.5.x libraries in these packages to make Chrome working again - Big smile

RubenKelevra commented on 2014-01-13 21:25 (UTC)

could you please just increase the version-number while adding a depency to the new version libgcrypt-1.6.0-1? That would fix it for everybody. :)

franziskus commented on 2014-01-13 21:03 (UTC)

@jz_train just reinstall the package, also gets you for a proper fix you have to wait for chromium/google to fix this

Det commented on 2014-01-13 20:16 (UTC)

And it's not out-of-date because of it..

Det commented on 2014-01-13 20:16 (UTC)

Or if you rebuild this thing.

jz_train commented on 2014-01-13 20:16 (UTC)

error while loading shared libraries: cannot open shared object file: No such file or directory libccrypt was updated today 12-16-14 to version libgcrypt-1.6.0-1-x86_64.pkg.tar.xz. If you downgrade to libgcrypt-1.5.3-1-x86_64.pkg.tar.xz chrome will run.

Det commented on 2014-01-09 19:45 (UTC)

Well, since I'm not the official bug tracker of Chromium, there's not much I can do. You should try the Arch forums or the actual bug tracker.

JoveYu commented on 2014-01-09 07:32 (UTC)

Hi, Det, remember my bug? I found my chrome render textinput without css to a input with 0px width, maybe this is the reason. i write <input type="text"> to a html file, and it display not really well!

Det commented on 2014-01-01 00:33 (UTC)

Even, if you redownload it? As in not using the one already in $startdir.

a1russell commented on 2013-12-30 07:44 (UTC)

eula_txt checksum is for me: 6d57da7476a4b1b7a81821d9c036425c

Det commented on 2013-12-26 13:56 (UTC)

Fixed. The package will now pull the previous from [core] (libgcrypt 1.5.3-1) and place it in /opt/google/chrome/. I decided to make this happen even, if libgcrypt 1.6.0-1 isn't currently installed on the system to allow the previous version to be installed _before_ upgrading to the new one.

Det commented on 2013-12-26 11:19 (UTC)


hadrons123 commented on 2013-12-26 11:17 (UTC)

gentoo has some fix for this post libgcrypt 1.6 update.

Det commented on 2013-12-26 11:16 (UTC)

Then how did you get in here? Liar.

hadrons123 commented on 2013-12-26 11:16 (UTC)

Doesn't start after new libgcrypt 1.6 update.

JoveYu commented on 2013-12-15 04:08 (UTC)

yeah nothing changed! same as chromium! version number is 31.0.1650.63 and my DE is cinnamon, everything is up-to-date!

Det commented on 2013-12-14 16:54 (UTC)

You sure? Cleared/renamed both ~/.cache/google-chrome/ and ~/.config/google-chrome/? The system itself is up-to-date? What's the version number in the "About" page?

JoveYu commented on 2013-12-14 08:00 (UTC)

I have tried, but nothing changed!

Det commented on 2013-12-13 08:56 (UTC)

Have you tried with a clean profile? Are you sure it's this channel and the thing is up-to-date?

JoveYu commented on 2013-12-12 01:19 (UTC)

why chrome render pages like this is it chrome's bug? or my computer's?

reed1 commented on 2013-12-11 04:41 (UTC)

@nhasian, not really for those who faces strange icon grouping issue with kde task manager, go here:

Det commented on 2013-12-07 22:31 (UTC)


tioammar commented on 2013-12-06 15:58 (UTC)

I got this error message.. ./opt/google/chrome/PepperFlash/ Lzma library error: Corrupted input data. Is that the problem with the pkg or my system?

nhasian commented on 2013-12-06 04:53 (UTC)

@reed1 I imagine some links were broken due to the name of the application changing from google-chrome to google-chrome-stable.

tarkin88 commented on 2013-12-04 20:04 (UTC)

Here a PKGBUILD, works for me Just change the source chrome link and md5sum

na12 commented on 2013-12-04 15:20 (UTC)

I get "[1204/] NaCl helper process running without a sandbox! Most likely you need to configure your SUID sandbox correctly" error and chrome then crash.

a1russell commented on 2013-12-04 01:36 (UTC)

eula_text checksum is failing.

reed1 commented on 2013-12-03 09:35 (UTC)

Anyone notice the shorcut isn't working anymore on KDE's task manager ?

magicalChicken commented on 2013-11-25 18:08 (UTC)

Could you please add libxtst to the dependencies, chrome crashes without it.

Det commented on 2013-11-25 08:44 (UTC)

Did you also notice the first comment from top?

mumbel commented on 2013-11-25 04:09 (UTC)

Does "google-chrome" exist in PATH currently? Right now I only have "google-chrome-stable" or is my PATH not including everything it should? Just noticed this when trying to set default browser.

Det commented on 2013-11-17 17:24 (UTC)

Just came here to say that now that the "Google Chrome Introduction Book" ( is becoming increasingly outdated the official Chromium blog is still here to keep you pretty damn interested in what they're doing to their project and the modern internet as a whole: I've also updated the Stable and Beta channel packages' .install files to also notify about the newly added -stable/-beta suffixes in the /usr/bin/ symlinks.

Commander commented on 2013-11-14 22:43 (UTC)

New version out: 31.0.1650.57

dmoebius commented on 2013-11-14 09:36 (UTC)

eula changed again. The new md5 sum is '6d57da7476a4b1b7a81821d9c036425c'.

ruario commented on 2013-10-23 08:19 (UTC)

@SysGhost: You have misunderstood. The checksum is not wrong for 30.0.1599.101. The reason it doesn't match the download is that the current stable Chrome is now 30.0.1599.114 (notice the higher stabilization branch number '114'). The problem is simply that the PKGBUILD is out of date, resulting in the version number and both the 32 and 64 bit checksums no longer matching. P.S. Read the PKGBUILD comments to understand how to check what the current version number is.

SysGhost commented on 2013-10-23 08:07 (UTC)

This PKGBUILD includes a 64 bit version of google-chrome too (Should be included in the "sources" section of this page) However, the md5sum in the PKGBUILD for the amd64 version "google-chrome-stable_30.0.1599.101_amd64.deb" is not correct. The correct md5sum should be: 7ff0eed996de3015683e0159b87d4d6e Replace the line that reads: [ "$CARCH" = 'x86_64' ] && md5sums[0]='9abdef7ee0422f7dabb16941453551cc' ...with: [ "$CARCH" = 'x86_64' ] && md5sums[0]='7ff0eed996de3015683e0159b87d4d6e'

Det commented on 2013-10-18 19:25 (UTC)

+1 yeah :D. If nothing else, glad to see new users acquainting themselves with PKGBUILDs.

petko10 commented on 2013-10-17 21:33 (UTC)

+1 for the a3a49767c80e92cca63b7a33aadfb2af for eula . I put that hash as the second md5sum (above the line [ "$CARCH" = 'x86_64' ] && md5sums[0]='e4ab342703de87ec117cfea3951acce8' ) , and the package installed normally.

grimsbain commented on 2013-10-17 17:28 (UTC)

After having an install error I downloaded the eula and checked the md5sum and it came back a3a49767c80e92cca63b7a33aadfb2af same as Cameri.

ichramm commented on 2013-10-17 16:28 (UTC)

Weird, just updated google-chrome and eula's MD5 is 7ae235c61533ba1108b03d42c49e00bd /cc @Cameri

Cameri commented on 2013-10-17 13:54 (UTC)

The eula changed. New md5sum: a3a49767c80e92cca63b7a33aadfb2af

beststone commented on 2013-10-17 07:40 (UTC)

I have same error. So I temporarily upgraded with "packer" command. $ packer -S --skipinteg google-chrome-beta Later fixed this package, it will upgraded normally.

mariusn commented on 2013-10-17 07:12 (UTC)

eula_text.html ... FAILED

joe1984 commented on 2013-10-16 22:00 (UTC)

I'm getting checksum a3a49767c80e92cca63b7a33aadfb2af for the file eula_text.html. Please fix this PKGBUILD.

MindfulMonk commented on 2013-10-03 11:34 (UTC)

Just change the PKGBUILD line [ "$CARCH" = 'x86_64' ] && md5sums[0]='e4ab342703de87ec117cfea3951acce8' to [ "$CARCH" = 'x86_64' ] && md5sums[0]='7f055de3dc3a62cedb5156420c1bc99a'

danielepaolella commented on 2013-10-02 09:33 (UTC)

On x86_64 I'm getting checksum: 7f055de3dc3a62cedb5156420c1bc99a

haaja commented on 2013-10-01 20:47 (UTC)

Just letting you know that there's an unnecessary .directory file in the tarball.

Det commented on 2013-09-20 18:26 (UTC)

Well, that's a bit longer than the line you already gave me anyway and the output is: <channel> <version> [...] Instead of: <channel>: <version> [...] And to be honest I don't really care about the "feeling" either, as long as it _is_ correct.

ruario commented on 2013-08-11 17:41 (UTC)

@Det: Just a thought, regarding your comment in the PKGBUILD that suggests using the rpm repo meta data to find out the latest version. Since you are now using the deb as the source perhaps this command (or a variant) might be more appropriate as it uses the deb repository metadata? Doesn't really matter given the version number is the same either way but perhaps it feels more correct. $ curl -s | gzip -d | awk -F "[ -]" '/Package: google-chrome/ { print $4; getline ; print $2 }'

Det commented on 2013-08-09 16:26 (UTC)

Yeah, I've noticed that too back when I made that decision.

zatricky commented on 2013-08-06 15:02 (UTC)

Minor non-bug: I've noticed the end-result is a .tar whereas most packages get packaged as .tar.xz

Det commented on 2013-08-02 15:09 (UTC)

I update it when I can.

lks commented on 2013-08-01 21:54 (UTC)

Save fast and stable browsers exists(See links in core). Google chrome has a nice user interface (and is not so save and sents all your data to google and NSA). Please add/ change your description. lks

workdowg commented on 2013-08-01 20:15 (UTC)

x64 new - md5sum 6d4367302896e62942184997770ddfa2

blackout24 commented on 2013-07-29 15:52 (UTC)

chrome://settings -> Advanced -> [ ] Enable Hardware Acceleration when available. To disable HW acceleration. It seems like it fixed it for me. No more zombie processes and freezes, when using NVIDIA.

oman002 commented on 2013-07-13 23:22 (UTC)

Chrome creates application shortcuts that look like this: /opt/google/chrome/chrome --app= But they don't work: $ /opt/google/chrome/chrome --app= /opt/google/chrome/chrome: error while loading shared libraries: cannot open shared object file: No such file or directory The following does work: $ /opt/google/chrome/google-chrome --app= Do I have to manually change my application shortcuts or can this be fixed in the PKGBUILD?

commented on 2013-07-11 06:39 (UTC)

Yeah its crazy it sometimes literaaly locks my computer up for about 1 minute. It seems the iowait percentage goes up to about 60% and than drops when it stops in resource monitor.

gadget3000 commented on 2013-07-09 14:27 (UTC)

I'm recently switched to using Chromium and I'm also getting weird freezes both on Chromium and Chrome. For me journald starts going mad whenever a freeze happens. I've submitted a bug to Chromium here:

TheWretched commented on 2013-07-08 14:48 (UTC)

I've had that sometimes (the random freezing). Moreso with chrome and it's stagnant processes than with chromium. I've had the occasional weird freeze with chromium too. Honestly chrome / chromium has become a major pain for me with the myriad of problems it seems to have ever since KDE 4.10, mostly because it's still the only decent browser despite all these problems, meaning one has to just live with it.

commented on 2013-07-08 05:08 (UTC)

Is it just me or does chrome or chromium just randomly lock up your machine for like 4 or 5 seconds randomly...... Not too sure what it maybe but it happens alot ..... It locks up only in chrome for about 5-10 seconds its very annoying ....

Det commented on 2013-07-03 22:41 (UTC)

You can just flag it.

fiveofakind commented on 2013-07-03 17:45 (UTC)

v71 md5sums: 03de683b0e9b8bf77db17060c8113d4d google-chrome-stable_current_amd64.deb 19ec74343b5396760d5d6c88d7282cfb google-chrome-stable_current_i386.deb

mike.cloaked commented on 2013-07-03 17:00 (UTC)

and hot on the heels of v70 now google-chrome-stable-28.0.1500.71-209842 is out!

Det commented on 2013-06-29 16:01 (UTC)

Added, but it'll be faster to just add that yourself. This just moves it to /tmp anyway so it's not a real fix.

JoveYu commented on 2013-06-28 11:22 (UTC)

chrome always create "libpeerconnection.log" in my HOME google say maybe you can fix it for the moment

Det commented on 2013-06-26 20:11 (UTC)

Yeah, I really wish our libarchive was built against lrzip: It's hugely faster than anything else and the compression is almost as good as with xz. If it actually became the default in Arch some day it'd be even better. I'm just not sure about the state of its downsides today.

mike.cloaked commented on 2013-06-26 16:29 (UTC)

Thank you for the change to .tar - it does save time to get the install done even if the file is a bit larger than the xz compressed one.

Det commented on 2013-06-26 10:32 (UTC)

Never ever heard of it (the software I mean). I mentioned this to e36freak.

mariusn commented on 2013-06-26 09:37 (UTC)

I'm using cower and meat

Det commented on 2013-06-26 09:36 (UTC)

I changed to PKGEXT='.pkg.tar' to speed up the compression that I'm so tired of waiting to finish. If you want it back you can just remove that line or add the ".xz" to the end. But I'm curious as to how did you guys actually encounter that error. Is it just another one of those AUR tools that don't support .pkg.tar's?

commented on 2013-06-26 08:24 (UTC)

@Det Seems to work like a charm - thanks ! :-) Besides the zombie processes that is. Even after a full update including nvidia drivers. So stupid ...

commented on 2013-06-26 08:13 (UTC)

Apparently it's the tar file google-chrome-28.0.1500.52-1-x86_64.pkg.tar. But it used to be a xz file indeed. Just run "pacman -U google-chrome-28.0.1500.52-1-x86_64.pkg.tar". At least, that works for me...

mariusn commented on 2013-06-26 07:40 (UTC)

==> Finished making: google-chrome 28.0.1500.52-1 (Wed Jun 26 10:39:54 EEST 2013) Installing `google-chrome'... error: error: no package file seems to have been created Any ideas?

mike.cloaked commented on 2013-06-24 08:55 (UTC)

In response to the comment on zombie processes by jan_goyvaerts - I also had occurrences of hanging chrome processes in version 27, but after I changed the setting under the advanced settings menu page, by unchecking "Continue running background apps when Google Chrome is closed" I have not had any hanging processes once I did that. However I still have not tried to modify the PKGBUILD file and update to v28 since I was not sure if there was a problem with building the new version on arch since this aur version was still not yet updated for a week.

commented on 2013-06-22 13:03 (UTC)

The url format has also changed to: "${_channel}_current_${_arch}.rpm"

commented on 2013-06-22 13:03 (UTC)

The url format has also changed to: "${_channel}_current_${_arch}}.rpm"

commented on 2013-06-22 13:02 (UTC)

The url format has also changed to: "${_channel}_current}${_arch}}.rpm"

donniezazen commented on 2013-06-20 15:10 (UTC)


commented on 2013-06-20 13:37 (UTC)

anyone else getting libpeerconnection.log in home folder now on 28.0.1500.45? that doesn't seem right...

Det commented on 2013-06-19 20:33 (UTC)

Pacman has lots of hacks and it destroys binaries.

commented on 2013-06-19 20:22 (UTC)

Why are there no zombie processes on other distro's ? Something in Arch has to interfere with Chrome...

commented on 2013-06-19 17:06 (UTC)

still get zombie processes on 28.0.1500.45-205727 with nvidia driver. back to nouveau driver :(

commented on 2013-06-19 14:27 (UTC)

The new version is 28.0.1500.45-205727 and the md5 sum for x86_64 is 7d2d46378af0f4a2e10a1ece957bd20a for anyone that wants to manually change the PKGBUILD manually until it gets changed.

ruario commented on 2013-06-18 08:31 (UTC)

@colinkeenan: Because there is a new version (28.0.1500.45-205727). Here is an easy way to find out what the current version is: curl -s | head -c96 | strings

colinkeenan commented on 2013-06-17 23:17 (UTC)

The download url doesn't work anymore (Monday June 17, 2013) ==> Retrieving sources... -> Downloading google-chrome-stable-27.0.1453.110-202711.x86_64.rpm... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 curl: (22) The requested URL returned error: 404 Not Found ==> ERROR: Failure while downloading google-chrome-stable-27.0.1453.110-202711.x86_64.rpm Aborting... ==> ERROR: Makepkg was unable to build google-chrome.

commented on 2013-06-16 19:03 (UTC)

A pity... because I need the nvidia driver for fast graphics. :-) Anyway, killall /opt/google/chrome/chrome also works.

commented on 2013-06-16 15:46 (UTC)

@jan_goyvaerts, sorry I was wrong about that. the issue is related to the nvidia driver. I switched to nouveau driver and have not had the issue going on 48 hours now.

Det commented on 2013-06-15 18:49 (UTC)

Well I'm not the one that's gonna test it.

commented on 2013-06-15 18:40 (UTC)

Replacing adblock by adblock+ didn't help ?

commented on 2013-06-15 17:16 (UTC)

switching from nvidia driver to nouveau driver fixed the zombie processes for me

Det commented on 2013-06-12 08:20 (UTC)


adityarajbhatt commented on 2013-06-10 16:30 (UTC)

I am using xfce4. chrome://version tells me I am using 27.0.1453.110 (Official Build 202711) Whenever I launch google-chrome from the terminal (to see details) I get the following errors : [741:766:0610/] Failed to get name owner. Got org.freedesktop.DBus.Error.NameHasNoOwner: Could not get owner of name 'org.chromium.Mtpd': no such name [741:766:0610/] Failed to get name owner. Got org.freedesktop.DBus.Error.NameHasNoOwner: Could not get owner of name 'org.chromium.Mtpd': no such name [741:741:0610/] Failed to call method: org.chromium.Mtpd.EnumerateStorages: object_path= /org/chromium/Mtpd: org.freedesktop.DBus.Error.ServiceUnknown: The name org.chromium.Mtpd was not provided by any .service files after which it works perfectly (everything - history, recently closed tabs and all). However it slows down the startup somewhat. I am just curious about how to remove these errors.

TheWretched commented on 2013-05-27 17:05 (UTC)

Spoke too soon. Still has stale processes that build up, still has tab crashes on middle click links. Version 27.0.1453.93-1 Going back to chromium.

TheWretched commented on 2013-05-27 16:24 (UTC)

Knock on wood, but so far no leftover and zombie processes on close, and no messed up tabs freezing on middle clicking links.

Zielony commented on 2013-05-25 17:10 (UTC)

If you have high-pitched audio using 27.0.1453.93 (like me), try to set default sample rate for pulseaudio daemon in /etc/pulse/daemon.conf like that: default-sample-rate = 48000 default-sample-format = s16le It helped for me. More:

commented on 2013-05-23 07:54 (UTC)

@t3ddy Impeccable as ever. Thanks ! :-)

mamamia88 commented on 2013-05-22 15:34 (UTC)

@mattijsf where did you find the verbld number?

netsurfer commented on 2013-05-22 13:40 (UTC)

@mattijsf i had to run "dos2unix PKGBUILD", then all ok! I dont now if its a my issue...probably not. thankyou

commented on 2013-05-22 06:53 (UTC)

google-chrome 27.0.1453.93 Updated PKGBUILD for the people who can't wait ;)

Det commented on 2013-05-22 05:26 (UTC)

I meant as in in here u cunt.

commented on 2013-05-22 05:05 (UTC)

Only the 104th ? Damn... :-p

Det commented on 2013-05-22 04:05 (UTC)

Yes there is and you are like the 104th person who's noticed it.

ccesar commented on 2013-05-22 03:52 (UTC)

There is a "always working" URL. I What is need then is only change the md5 the new release.

mamamia88 commented on 2013-05-21 21:13 (UTC)

@t3ddy i made some modifications to your pkgbuild that you may want to take under consideration. I basically just changed the verbld line to current and the sources line to use the direct download link from google. All you would have to do is change is the pkgver and md5sum lines with each release. Here is a link to check out if you like. On a side note I don't even think that build 27 has hit the google servers yet at least for linux since I just installed with my modified pkgbld and it still says version 26 in about chrome

mamamia88 commented on 2013-05-21 20:14 (UTC)

@jan_goyvaerts apparently it has been released but in order to update the maintainer needs to know the correct verbld number which i can't for the life of me find on because it's loading so slowly. md5sum for new build are: 29fa12a2495e35b15fb384523b6535e0(i386) and 741603ac7c4cf65a62ffe73c55c18351(64bit). You just need to find that one little piece of info before upgrading manually I think but like I said they must be getting hammered right now so it's hard to find

commented on 2013-05-21 18:54 (UTC)

aha ! chrome 27 has been released ?

weedfreak commented on 2013-05-19 09:35 (UTC)

get error error: google-chrome: missing required signature :: File /var/cache/pacman/pkg/google-chrome-26.0.1410.63-1-x86_64.pkg.tar.xz is corrupted (invalid or corrupted package (PGP signature)).

Det commented on 2013-05-18 23:20 (UTC)

I see you are also experiencing vision problems.

latelx commented on 2013-05-17 05:55 (UTC)

experice flash audio problem.

Det commented on 2013-05-14 13:15 (UTC)

No, you don't have to email him, because he's already getting notifications of our comments.

commented on 2013-05-13 22:32 (UTC)

Thanks @carukia & @Det, I've sent an email to the maintainer.

Det commented on 2013-05-13 17:58 (UTC)

@carukia, and I know it. See:,

commented on 2013-05-13 17:26 (UTC)

@ryanpcmcquen I think only the maintainer can remove it.

commented on 2013-05-13 16:51 (UTC)

Oops, accidentally flagged this out of date! Does anyone know how to remove the flag?

stuntgp2000 commented on 2013-05-10 13:16 (UTC)

for those who suffer from sound glitches when playing flash videos they should launch google-chrome with this parameter --audio-buffer-size=2048 $ google-chrome --audio-buffer-size=2048 this issue comes from upstream and it affects v26 and v27 beta

commented on 2013-05-10 06:47 (UTC)

Processes are still remaining into memory. Shame...

commented on 2013-05-08 19:47 (UTC)

New libpng is pushed, resulting in this: $ google-chrome /usr/bin/google-chrome: error while loading shared libraries: cannot open shared object file: No such file or directory

michalmiddleton commented on 2013-04-26 19:12 (UTC)

There is a "choppy audio" bug in the version 26 (it think it was introduced together with new the new flash 11.7). The workaround is to start google-chrome with "--audio-buffer-size=2048" parameter. Optionally, you can edit the Chrome shortcut (/usr/share/applications/google-chrome.desktop) and change the line "Exec" to: Exec=/opt/google/chrome/google-chrome --audio-buffer-size=2048 %U

sjakub commented on 2013-04-17 06:37 (UTC)

TheWretched: I have the exact same issue. Both with chrome (all channels) and chromium. I think it happens way more often with chrome though.

TheWretched commented on 2013-04-11 15:18 (UTC)

Aaaaand as fast as I typed that and went and tested it, chromium also does the middle click broken tab freeze for me as well. I have seen no one else comment on this anywhere I've searched, so if anyone has seen this or knows anything at all about it I would be very grateful to hear about it.

commented on 2013-04-11 15:17 (UTC)

You'll get the tray icon by checking "Continue running background apps when Google Chrome is closed" in the settings. I'm not sure, but when this is enabled chrome won't leak processes any more. Each time chrome is restarted the number of processes is not higher as the last time. Which is the case when this option is not set. Each time more and more processes are added. Unfortunately selecting "exit" on the tray icon does *not* stop the processes. Still needed to do killall chrome. :-/

TheWretched commented on 2013-04-11 15:14 (UTC)

Going to give chromium another go. When I first upgraded to 4.10 I was having problems with it too. I don't think this is related to the stale processes but I have another major annoyance in chrome that I was also getting in chromium with 4.10. That is when I middle click a link to open in a new tab it will frequently freeze the browser and result in a tab that looks like about a quarter the size it should be. Again resulting in a pkill being needed. This persists even with the latest KDE and the latest chrome as of this writing. I guess well see if this is still the case with chromium.

Det commented on 2013-04-11 12:49 (UTC)

They like your memory. They don't wanna go. You should be flattered.

commented on 2013-04-11 07:46 (UTC)

Processes are stuck in memory again. Damn...

Det commented on 2013-04-09 18:40 (UTC)

I don't.

MrTux commented on 2013-04-09 18:10 (UTC)

source=("${_channel}_current_${_arch}.rpm") I think use this is better for update

commented on 2013-04-08 21:47 (UTC)

FYI: If you are having problems with zombie processes (Chrome_ProcessL, etc.) and you are using the nvidia blob driver, the problem is known to upstream ( and is in fact due chrome's tcmalloc clashing with a bug in the nvidia driver. The only "workaround" is to use chromium from [extra], AFAIK, since it disables tcmalloc at compile time.

Det commented on 2013-04-04 10:55 (UTC)

Let's hope 4.10.2 makes things work again. Seems like the whole 4.10 series was a bit half-assed anyway:

commented on 2013-04-04 06:48 (UTC)

Yesterday I had a Chrome tray icon - with which you can tell the thing *not* to remain in background when closed. This actually works for me ! :-) So they have been messing with processes...

heaven commented on 2013-04-03 19:45 (UTC)

Also have all those problems with sound and leftover processes.

gabrielrcp commented on 2013-04-02 04:09 (UTC)

ttf-font is provided by ttf-ms-fonts (amongst other true type fonts) Relevant discussion:

commented on 2013-04-02 02:38 (UTC)

Hi: 'ttf-font' package does no exist, so use 'ttf-ms-font' instead and correct PKGBUILD please. Thanks.

TheWretched commented on 2013-04-01 17:36 (UTC)

Yeah it has been awful lately. When I upgraded to KDE 4.10 + Xorg 1.14 it became disastrous. It's gotten slightly better but still I have to constantly kill it's leftover and zombie processes that build up over time. I went so far as to test it on a fresh user account with the same results. Just awful. So bad I've even resorted to using Firefox a lot more.

kiodo1981 commented on 2013-03-31 13:32 (UTC)

Audio problem. I solved starting with google-chrome %U --audio-buffer-size=2048

commented on 2013-03-29 11:46 (UTC)

Thanks ! It builds and seems to run properly. I hope the cleanup/repackaging/... for release 26 is the same as for 25. :-)

Det commented on 2013-03-29 08:42 (UTC)

Just use either of these: source=("${_channel}_current_${_arch}.rpm") or (debs: quicker) elif [ "$CARCH" = "x86_64" ]; then _arch='amd64' [...] source=("${_channel}_current_${_arch}.deb") ..and build with "--skipinteg".

commented on 2013-03-29 08:29 (UTC)

Newbie question: can the current make script be updated with the source & md5 indicated by k2s ?

commented on 2013-03-28 08:53 (UTC)

Don't know what happens lately, but Chrome is leaking memory big time ! Hope this release resolves this problem.

k2s commented on 2013-03-27 12:33 (UTC)

current 64bit: source=('') md5sums=('76524edb4ce9d3f946c050a48e831437')

Det commented on 2013-03-27 09:26 (UTC)


commented on 2013-03-27 09:25 (UTC)

Is it possible there was an update recently ?

pwalker commented on 2013-03-22 13:50 (UTC)

I'm not quite sure how to contribute to this, but can you add ttf-google-webfonts in the the optdepends? I know the web will look alright in chrome without it, but I've noticed a few style issues on some sites like github, banks, etc. that were solved by having the appropriate fonts. Checkout for a comparison.

Det commented on 2013-03-14 12:35 (UTC)

Uhhh.. you're asking how can the _current_ url get the pkgver the same way the versioned one does? Well, you find it out first the same way it's done now, silly. What did you think the current solution is some ungreppable magic that automatically updates itself :D? If you're asking what to do with the local one that's already there (since it's gonna have the same name), then you do a: source=("google-chrome-${_channel}_${pkgver}_${_arch}.rpm"::"http(s)://${_channel}_current_${_arch}.rpm") Which has the only downside that it's very fugly and renders the Source text up there to just "google-chrome-${_channel}_${pkgver}_${_arch}.rpm", as in with google-earth (still points to the real file, though).

commented on 2013-03-14 12:16 (UTC)

@Det: Thanks! But what I meant is how does a script know that it has to update a local Google Chrome version to a newer one from the internet? It can't read that blog...

Det commented on 2013-03-14 00:48 (UTC)


commented on 2013-03-13 18:56 (UTC)

@mdyan: Agree. @XenGi: How do you know what version you download exactly when you get something from "...current..."?

mdyan commented on 2013-03-13 17:25 (UTC)

I don't mean to speak for the package maintainer, but I think that it's for versioning purposes.

XenGi commented on 2013-03-13 14:11 (UTC)

Why not using the link to the latest stable instead of building the url from some blog posting?$arch.rpm Where $arch is x86_64 or i386.

commented on 2013-03-13 09:43 (UTC) seems to be invalid, too :(

Det commented on 2013-03-11 17:58 (UTC)

I think it was due to Xorg 1.14.

TheWretched commented on 2013-03-11 17:53 (UTC)

For what it's worth, version 25.0.1364.160-1 seems to work a lot better on kde 4.10 for me having updated to it this morning. I haven't got the middle click = messed up tab + browser freezing issue so far. Knock on wood.

mfs commented on 2013-03-08 10:29 (UTC)

The stable_current_x86_64 is already 25.0.1364.160.

commented on 2013-03-08 03:18 (UTC)

The source link "" is invalid. Use the following URL instead?

leduck commented on 2013-03-01 18:23 (UTC)

You can add the following line in the PKGBUILD file? sed -i '1,/^$/{/^$/s/^$/Keywords=Internet;WWW;Browser;Web;Explorer;\n/;}' /usr/share/applications/google-chrome.desktop It serves to add the tags to search the gnome. Thanks

TheWretched commented on 2013-03-01 16:12 (UTC)

I thought flash was working for me, I know it is for sure in chromium. The actual flashplugin that is not the pepper flash since chromium doesn't have that. But I was just in the plugins in chromium and know they work there. I've resigned to switch to chromium for now on account of 4.10

12eason commented on 2013-03-01 14:23 (UTC)

flash seems to be broken under 4.10 too. In fact, there's nothing listed under chrome://plugins/

bred commented on 2013-02-28 21:54 (UTC)

@ TheWretched I agree this version seems to bee buggy under KDE 4.10.

TheWretched commented on 2013-02-28 20:09 (UTC)

Is anyone having issues with this? I recently updated to KDE 4.10 and decided to update chrome as well. I built this version (25.0.1364.97-1) manually, pulled off the old, removed the settings and cache folders from my home folder and loaded up this version. Problem I am having is after closing there are a bunch of processes left running from chrome, and it behaves sporadically, most notable issue, middle clicking a link for a new tab results in a broken tab that is about 1/8 as wide as a normal tab and the browser freezes. Closing all the remaining processes manually and restarting temporarily resolves but problem recurs. Not experiencing this problem with Chromium current version.

Det commented on 2013-02-24 16:48 (UTC)

You mean pacman -R libcanberra-pulse?

commented on 2013-02-24 16:45 (UTC)

For those having issues with Chrome or Chromium freezing when something is typed in the address bar, try the fix outlined here:

Det commented on 2013-02-18 10:25 (UTC)

Sounds like corrupt settings or something. You tried with a clean profile and syncing your settings or so?

mityukov commented on 2013-02-18 08:23 (UTC)

There is one strange issue (that started several updates before): the browser simply closes when I type something into the addressbar (not every time, though). Without notices, etc. Note: I have the following in the dmesg, but not sure if it's google-chrome related: > [ 2897.267727] traps: chrome[3233] trap divide error ip:7f3e46775170 sp:7fff73cf4c60 error:0 in chrome[7f3e45dda000+4c45000]

Det commented on 2013-02-12 20:37 (UTC)

You can just flag it when the URL becomes 404, as t3ddy already advised you to. It just means there's a new version and the old one (whether it be a symlink or the actual file) was removed from the server.

commented on 2013-02-12 20:19 (UTC)

Please fix URL!

ckoch786 commented on 2013-02-12 16:28 (UTC)

Thank you for your contributions!

commented on 2013-02-07 08:57 (UTC)

Works - thanks a lot for keeping it up to date ! :-)

t3ddy commented on 2013-02-06 21:14 (UTC)

When you get 404, just flag the package out-of-date and I'll update it ;)

commented on 2013-02-06 20:17 (UTC)

It has been fixed, at least it doesn't throw a 404 anymore for me.

ingamedeo commented on 2013-02-06 19:50 (UTC)

Yes....404 error here! Please fix it!!

nickoe commented on 2013-02-06 19:43 (UTC)

Download link does currently throw a 404.

Det commented on 2013-02-06 18:14 (UTC)

No, it was kind of updated in between whenever it was.

mityukov commented on 2013-02-06 18:12 (UTC)

The download link was reported as broken yet in January, and it is still in this state. I can use links, provided by leduck, but I'm wondering if the package is still supported?

dangersalad commented on 2013-02-06 18:11 (UTC)

I changed the url to to get it to work. The one provided in the comment on 01-31-2013 no longer is working (for me)

commented on 2013-02-06 09:16 (UTC)

Worked like a charm. Thanks !

commented on 2013-02-06 09:04 (UTC)

Worked like a charm. Thanks !

commented on 2013-02-04 20:10 (UTC)

404 for me. bad link.

vendion commented on 2013-02-04 19:34 (UTC)

I seem to be getting 404's off the URLs provided in the PKGBUILD, another update maybe?

leduck commented on 2013-01-31 00:18 (UTC)

New URL for i386 MD5SUM = 2AE939E664AE538879F56D7C3C8C9970 New URL for x86_64 MD5SUM = 6E374CBEB5314290E9F0C6586798B2E9

commented on 2013-01-31 00:12 (UTC)

Had to modify the url in the PKGBUILD

Det commented on 2013-01-30 23:43 (UTC)


TheWretched commented on 2013-01-30 22:07 (UTC)

Package is broken, download source no longer exists with google.

jskier commented on 2013-01-28 20:54 (UTC)

I've been having issues with Chrome recently:

Det commented on 2013-01-26 20:36 (UTC)

I had that for a long time but then it just went away. It was caused by the history file getting corrupted. I'm assuming they fixed it but not those already corrupt files.

mityukov commented on 2013-01-26 20:20 (UTC)

It crashes few times a day when starting typing anything in the addressbar (live search is enabled). Does anyone else have this issue? Note: I haven't got it in Ubuntu, but I get it here, although it was updated two or three times already (so it's not certain-version-specific).

ruario commented on 2012-12-12 08:22 (UTC)

@codygman: No the URL scheme has not changed, the current (as I write this) latest build can be downloaded directly here with the version and revision numbers still present: The URL you pointing to works like a symlink to the actual build.

commented on 2012-12-12 03:32 (UTC)

url scheme changed.. new url:

commented on 2012-12-11 01:46 (UTC)

bad md5sums! I corrected it then I got another error: -> Extracting google-chrome-stable-23.0.1271.95-169798.x86_64.rpm with bsdtar ./opt/google/chrome/PepperFlash/ truncated bzip2 input

dejavu commented on 2012-12-06 12:59 (UTC)

@Qoru: Thanks, enabling the switch "Override software rendering list" did the trick. Seems that my nvidia card was blacklisted recently.

commented on 2012-12-03 14:17 (UTC)

Is it possible the older versions of chrome can not be downloaded from any more ?

mike.cloaked commented on 2012-11-30 09:52 (UTC)

It works nicely for me for version 23.0.1271.95-1

mike.cloaked commented on 2012-11-29 21:22 (UTC)

Another new version released today (29th November 2012)

commented on 2012-11-27 19:11 (UTC)

Qoru's solution worked with Nvidia for me.

Det commented on 2012-11-27 18:41 (UTC)

Not with me it is.

commented on 2012-11-27 18:30 (UTC)

Flash videos are playing but sound is distorted in this version.

commented on 2012-11-23 19:58 (UTC)

To solve the full screen issue: Go to chrome://flags, then enable "Override software rendering list". This worked for me with the radeon open source driver. Video is also much faster now. I also changed some other settings, like enable WebGL (works fine), etc., but the item mentioned above did it. It worked in Arch 32-bit and 64-bit for me. Other settings which may or may not be related: $ cat /etc/X11/xorg.conf.d/20-radeon.conf Section "Device" Identifier "Radeon" Driver "radeon" Option "AGPMode" "8" Option "ColorTiling" "on" Option "AccelMethod" "EXA" EndSection

dejavu commented on 2012-11-19 13:21 (UTC)

Sorry i meant 'software video rendering'. Previously I got 'accelerated video rendering' on my Nvidia card. On my ATI card, using the catalyst driver the acceleration works though. Independently the packager can't do anything about this. ;-)

dejavu commented on 2012-11-19 00:32 (UTC)

Flash doesn't work well in this release, using Nvidia card with proprietary driver. Now I have only 'software video decoding' instead of hardware (Youtube). Also the tab title isn't shown all the time.

Det commented on 2012-11-14 11:04 (UTC)

Like the Intel/Catalyst drivers or some special kernel.

Willrandship commented on 2012-11-14 07:26 (UTC)

Well, I can confirm that fullscreen flash IS it's not a general problem. It's something on your end.

rumpelsepp commented on 2012-11-12 17:46 (UTC)

In addition to the problem with youtube videos, there is a problem with page titles on the tabs (html tag: <title>). For example a tab with opened Gmail only shows the favicon, but no text.

Det commented on 2012-11-10 19:08 (UTC)

Like f*ck it doesn't.

commented on 2012-11-10 18:55 (UTC)

Confirmed, fullscreen flash video on youtube does not work with the latest update

stdigger commented on 2012-11-07 12:34 (UTC)

not work fullscreen flash video =(

Det commented on 2012-10-07 20:17 (UTC)

Yes :D. But it won't use it.

msx commented on 2012-09-26 00:13 (UTC)

I agree that in a public packages one should try to do things in a sane way whenever is posible but sometimes the only way to deal with real life problems is with ugly hacks (and they're lovely): it isn't bad at all to get your hands dirty sometimes... in fact it's very fun and hackish!

commented on 2012-09-09 22:13 (UTC)

Thanks! I figured it was a horrible idea, but it made it work for now. I certainly don't recommend it. Why was it required in the first place? Is there a better solution?

Det commented on 2012-09-09 22:09 (UTC)

You have a worse judgment! >:O

commented on 2012-09-09 22:07 (UTC)

Hello! Although a symlink was created for in /opt/google/chrome, I don't think chrome is looking for libraries there (perhaps you would have to set the "run in" or the like to that directory?) I put a sym link in /usr/lib against my better judgement to get it to work. Is there a better work around or did I mess something up on install?

Vrtak-CZ commented on 2012-08-14 23:45 (UTC)

Det commented on 2012-08-14 18:23 (UTC)

That's why I flagged it.

mike.cloaked commented on 2012-08-14 18:18 (UTC)

... yet another new version - .79 is out now.

ruario commented on 2012-08-14 10:44 (UTC)

@Det: Ok, well as I said I am not personally adverse to using the deb again if t3ddy wants to save some bandwidth for users of this PKGBUILD. I honestly thought that Google would have LZMA or XZ rpms available by now and as you know I suggested the format of rpms as they are more consistent with Arch conventions, making for a potentially easier to maintain PKGBUILD. I was *trying* to help, not inconvenience people. ;) Anyway, I suspect the bigger plus for t3ddy in switching to rpm was a predictable way to gather the ${_verbld} information for the meta script he uses to update the PKGBUILD. The command I supplied him previously did this. However, given that I now see that you can also query the APT repository to gather this information (as outlined in my previous post) I guess that advantage is moot. P.S. I guess I am lucky that my internet connection is fast enough that even with a 23% increase I can still download the rpms within a couple of seconds.

Det commented on 2012-08-14 10:01 (UTC)

Actually more than slightly. The 64-bit .77 rpm is 42.2MB, while the deb is only 32.3MB. That's a -23% difference.

ruario commented on 2012-08-14 07:55 (UTC)

@tancrackers: No it isn't 'optimised' for Ubuntu. The package contents of the rpm and deb files are identical apart from a different updater cron job file and 3 extra Debian specific files in the deb. You can confirm this as follows: $ mkdir rpm deb $ wget -qO- | bsdtar -xf- -C rpm $ wget -qO- | bsdtar -xOf- data.tar.* | bsdtar -xf- -C deb $ diff -qr rpm deb Files rpm/etc/cron.daily/google-chrome and deb/etc/cron.daily/google-chrome differ Only in deb/usr/share: applications Only in deb/usr/share: doc Only in deb/usr/share: menu I would imagine that Google build the binaries only once per architecture and then package them into to different formats (this is also what happens with Opera). @Det: If t3ddy wants to use the debs as a source, it is fine by me. I suggested the rpms as source because because they simplify the packaging steps, i.e. they are auto-unpacked by makepkg, they are more generic (less distro specific stuff to remove) and because at the time I felt it was easier to automate gathering the version and revision (${_verbld}) information, allowing for source URLs that reflect the actual build being packaged. As it happens I since discovered you can gather this information by parsing the APT repository meta-data. As you state however the deb files are slightly smaller, though I was hopeful that by now Google would have fixed this. There is nothing stopping them using LZMA/XZ compressed rpms as they can be handled by all their supported rpm-based distros (currently only Fedora and openSUSE). Anyway, for those wishing to manually update the PKGBUILD, you can work out the latest version and revision (${_verbld}) information (to update the URL) by reading it out of the first 96 bytes of data in the header of the rpm, like so: $ wget -qO- | head -c96 | strings | rev | awk -F"[:-]" '/emorhc/ { print $1 "-" $2 }' | rev 21.0.1180.77-150576 Alternatively you can pull it out of the YUM repository meta-data: $ wget -qO- | gzip -d | awk -F\" '/-stable/ { print $10 "-" $12 }' 21.0.1180.77-150576 Or if you are using the .deb as a source from the APT repository meta-data: $ wget -qO- | gzip -d | awk '/Package: google-chrome-stable/ { getline ; print $2 }' 21.0.1180.77-r150576 The format of the version specific URLs is as follows:${_arch}/google-chrome-${_channel}-${_verbld}.${_arch}.rpm or${_channel}/google-chrome-${_channel}_${_verbld}_${_arch}.deb

mike.cloaked commented on 2012-08-13 21:02 (UTC)

Seems that version 21.0.1180.77 was pushed out rather soon after the previous version!

Det commented on 2012-08-12 10:37 (UTC)

Naturally. Deb: Rpm:

antihero commented on 2012-08-12 09:57 (UTC)

Det do you have an up-to-date PKGBUILD I could use?

Det commented on 2012-08-12 09:55 (UTC)

Because the guy I had this discussion with felt .rpm's were more fit for Arch due to its lack of debianisms (the .menu and changelog.gz) even after it was revealed they use an inferior compression algorithm (Bzip2 vs LZMA). E: We originally used debs. I still do with my local package.

tancrackers commented on 2012-08-12 02:02 (UTC)

Why not use the .deb packages instead? Isn't chrome optimized mostly for Ubuntu anyways?

Det commented on 2012-08-11 16:17 (UTC)

Yeah, as already stated.

crimsonknave commented on 2012-08-11 14:57 (UTC)

DL URL that I could find that worked was which worked with the checksum forbidden404 posted (for x86_64)

Det commented on 2012-08-09 16:36 (UTC)

It changes on every new build. That's just the general one to which I assume the latest one is linked to.

commented on 2012-08-09 15:23 (UTC)

The url changed, I've edited the PKGBUILD if [ "$CARCH" = "i686" ]; then _arch='i386' md5sums=('860ac2e244b5e096f3896bc05109cda6 ') elif [ "$CARCH" = "x86_64" ]; then _arch='x86_64' md5sums=('1c9b94721563eaf3e6da03978597c404') fi source=("${_channel}_current_${_arch}.rpm")

commented on 2012-08-01 01:38 (UTC)

I tried to update my chrome today and it was out of date. What happened? All the Google Chrome versions are marked as out of date.

Det commented on 2012-07-22 11:47 (UTC)

And so you figure 10 bytes is the way to go then? If Chrome uses too much space, you can clear your cache more often. If you actually want to limit its usage you can do that as well, but I wouldn't appreciate having that shoved into the package by default. E: Well, actually I don't even use the .desktop so as long as these things stay away from the launcher/symlink in /usr/bin, I don't really care.

gratuxri commented on 2012-07-21 16:50 (UTC)

Because of Disk Quota, sometimes need chrome over 500 Mb

aliasbody commented on 2012-07-14 22:35 (UTC)

Just a silly question... Why isn't this package on the [extra]'s repositorie ?

t3ddy commented on 2012-07-10 14:36 (UTC)

Ok, I've read the meaning of that flag, but I don't understand why someone should set it in that way.

Det commented on 2012-07-10 14:08 (UTC)

He wants a disk cache of 10 bytes.

t3ddy commented on 2012-07-10 13:41 (UTC)

May I ask you why?

gratuxri commented on 2012-07-10 12:05 (UTC)

I think, that it was be usefull if package-maintainer would change the .desktop file, so that chrome starts with --disk-cache-size=10 option or similar.

sjakub commented on 2012-07-04 15:47 (UTC)

Could it link to /usr/lib/ instead? 1.0.1 is installed by systemd-tools in [core], but the version in testing installs 1.0.2 instead. (or even would work on both.

ataraxia commented on 2012-06-28 22:52 (UTC)

This doesn't need to depend on libpng12 any longer, it's been updated to use the current libpng (1.5).

commented on 2012-06-28 18:58 (UTC)

True that! Disabling pepperflash did the trick (fyi to disable it go to: chrome://plugins/ click details on the top right and disable pepperflash from there). Wouldn't it be useful having it working though?

konart commented on 2012-06-28 17:28 (UTC)

Okay, the problem was in Shockwave Flash 11.3 r31 /opt/google/chrome/PepperFlash/ after turning pepper flash off - everything back to normal

konart commented on 2012-06-28 17:17 (UTC)

Also chrome spams with those messages: ALSA lib pcm_direct.c:980:(snd1_pcm_direct_initialize_slave) unable to install hw params ALSA lib pcm_dsnoop.c:623:(snd_pcm_dsnoop_open) unable to initialize slave

konart commented on 2012-06-28 17:13 (UTC)

Don't know about youtube, however in my case the browsing itself is almost impossible. Chrome lags like hell - scrolling, input, everything. It's like a slideshow

commented on 2012-06-28 13:05 (UTC)

Also having issues with fullscreen mode in youtube. When clicking the fullscreen button the image only flickers a short time and the video stays the same size. The new version of chromium works flawlessly.

commented on 2012-06-28 12:48 (UTC)

Anyone else having issues playing fullscreen youtube videos after the last update?

Fumon commented on 2012-06-02 23:25 (UTC)

On a new install done the day of this post, xfce google-chrome would report "Aborted" Installed chromium which only pulled "ttf-dejavu" google-chrome would then work. I'm assuming the dependencies need to be updated with a font.

martadinata666 commented on 2012-06-02 18:25 (UTC)

thanks @eric chrome work flawlessly again

t3ddy commented on 2012-06-02 14:07 (UTC)

I've added the workaround to the pkgbuild. You can remove the simlink to keep the system cleaner.

commented on 2012-06-02 13:37 (UTC)

Thanks ericcoleman

msx commented on 2012-06-02 08:06 (UTC)

@ericcoleman: +1

commented on 2012-06-01 23:16 (UTC)

Temporary workaround for jstoik1 and vanni971's problem (as well as others I'm sure) $ sudo ln -s /usr/lib/ /usr/lib/

commented on 2012-06-01 21:54 (UTC)

Same problem of jstoik1

jstoik1 commented on 2012-06-01 18:36 (UTC)

Following the switch from udev to systemd-tools, google-chrome fails to run: error while loading shared libraries: cannot open shared object file: No such file or directory

donniezazen commented on 2012-05-31 21:47 (UTC)

Is flashplayer included in this build of chrome or inbuilt flash is only included with dev channel?

nDray commented on 2012-05-16 12:47 (UTC)

More precisely 19.0.1084.46

commented on 2012-05-16 12:21 (UTC)

Chrome 19 is out!

commented on 2012-05-05 19:21 (UTC)

Well I am not 64 bit :)

Det commented on 2012-05-01 11:36 (UTC)

*Scratch that, Skype's there (and so were my jre and jdk packages)* I dunno. I suppose, if re-packaging Skype is okay, then there should be no issues in doing so with Google Chrome or even Google Earth: But even if they did move this some day, I'm not sure they'd keep both.

commented on 2012-05-01 06:40 (UTC)

I think google chrome should be in the extra repository. I don't like Chromium.

Det commented on 2012-04-29 11:19 (UTC)

Because you are so 64-bit.

commented on 2012-04-29 08:50 (UTC)

Why I don't have flash inbuilt for google chrome ?

alexcortes commented on 2012-04-13 16:31 (UTC)

Not working here without libpng12.

commented on 2012-04-07 08:46 (UTC)

Thanks :)

t3ddy commented on 2012-04-07 08:33 (UTC)

fixed :)

commented on 2012-04-07 08:10 (UTC)

I removed libpng12 get error /usr/bin/google-chrome: error while loading shared libraries: cannot open shared object file: No such file or directory.

commented on 2012-04-07 01:23 (UTC)

I'm getting a 404 error.

mike.cloaked commented on 2012-04-06 16:19 (UTC)

Any chance this can be updated to version 18.0.1025.151? Thanks

nDray commented on 2012-03-29 12:58 (UTC)

It's a flash problem. See this, for example:

cowai commented on 2012-03-29 12:55 (UTC)

Anybody experiencing blueish overlay on flash videos with the latest chrome? All youtube looks like Avatar now :P

Det commented on 2012-03-28 18:14 (UTC)

Yeah. That usually happens when they update their software and remove the old one.

Svenw commented on 2012-03-28 17:53 (UTC)

Looks like google has changed their download path as I get a 404 error on the AUR install, directly using the link above, or browsing the google site.

t3ddy commented on 2012-03-22 16:28 (UTC)

my mistake, again :)

wellington commented on 2012-03-22 14:58 (UTC)

Don't running without libpng12

commented on 2012-03-22 02:06 (UTC)

17.0.963.83 is available now.

Det commented on 2012-03-12 12:36 (UTC)

What xduugu meant was that this was _supposed_ to depend on 'openssl098' (0.9.8.t) - instead of the even older 'openssl-compatibility' (0.9.8.n) that he put there by mistake.

kkimdev commented on 2012-03-12 01:50 (UTC)

@t3ddy wait, your PKGBUILD still depends on openssl098, not openssl-compatibility.

t3ddy commented on 2012-03-11 12:54 (UTC)

Fixed, thanks.

xduugu commented on 2012-03-11 09:27 (UTC)

I just want to point out that this new release depends on openssl-compatibility again. Dunno if this was intended.

t3ddy commented on 2012-03-11 08:50 (UTC)

This is not a good news since I'm using the same template for the three versions of chrome.

sanusart commented on 2012-03-11 08:33 (UTC)

[ Temporary ] Working PKGBUILD:

Det commented on 2012-03-10 19:43 (UTC)

T3ddy just figured that if the Dev Chrome was using it, so should this.

commented on 2012-03-10 13:16 (UTC)

libpng12[0] is still a dependency, it seems to have gone missing from the PKGBUILD. [0]:

Det commented on 2012-03-09 14:02 (UTC)

Because it's out of date.

crabman commented on 2012-03-09 13:55 (UTC)

building fails with: curl: (22) The requested URL returned error: 404

Det commented on 2012-03-08 20:35 (UTC)

Yes. Lucid doesn't use it.

mike.cloaked commented on 2012-03-08 20:06 (UTC)

Is there any reason it can't depend instead on the openssl 1.0 from core?

xduugu commented on 2012-03-08 14:31 (UTC)

Could you please change the openssl-0.9.8 dependency from openssl-compatibility[1] to openssl098[2], since the former uses a build from 2010 which contains security issues. Thanks. [1] [2]

Det commented on 2012-03-07 18:56 (UTC)

Uhhhhhhh... no?

mike.cloaked commented on 2012-03-07 18:49 (UTC)

By the way libpng12 on which chrome depends was updated on Feb 20th in the AUR - in order to be up to date this package should also be updated.

mike.cloaked commented on 2012-03-07 18:41 (UTC)

Everything seems to work OK for me installing the new version .....66 though I am doing this some hours after the date/time on the previous comments - I had previously installed openssl-compatibility from the AUR. If you tried to install openssl-compatibility from the normal pacman repos then it won't work.

BasioMeusPuga commented on 2012-03-07 08:40 (UTC)

I'm getting the same 404 error. The URLs are out of date. The following works, however: 1. Put the md5sum of the chrome rpm in the pkgbuild ("388db19d984d7cfabac98ec2364b6d42" for x86_64) 2. Use the following URL in the source: "" Seems to build just fine after that.

commented on 2012-03-07 07:19 (UTC)

I have this error: -> Downloading google-chrome-stable-17.0.963.65-124586.x86_64.rpm... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 curl: (22) The requested URL returned error: 404

sanusart commented on 2012-03-07 06:01 (UTC)

Just a quick one

commented on 2012-03-07 02:41 (UTC)

@jarav spotify installed those.

Hexcles commented on 2012-03-07 00:18 (UTC)

Package out of date. File not found(404) on

commented on 2012-03-06 18:57 (UTC)

I am having problems because of the openssl-compatibility dependency. I am not able to install that package because of errors: error: failed to commit transaction (conflicting files) openssl-compatibility: /usr/lib/ exists in filesystem openssl-compatibility: /usr/lib/ exists in filesystem Errors occurred, no packages were upgraded.

mike.cloaked commented on 2012-03-05 22:43 (UTC)

OK I have just seen the other comments from early January now - I did not spot that there were additional comments not listed by default on this page unless specifically clicking on the appropriate link - apologies as I am fairly newly converted to arch from Fedora which I was using for the past 8 years! I note that there were similar suggestions but not exactly the same as mine early on in the discussion from a couple of months back - anyway if t3ddy can utilise the additional info from my earlier comment in his script then it may be of help to him in maintaining chrome updates. It does need a little scripting work but it should be do-able. The only change in the PKGBUILD file would then be a switch to sha1sum instead of md5sum but that is a minor change.

mike.cloaked commented on 2012-03-05 22:18 (UTC)

OK - no problem - if the t3ddy script that generates the pkgbuild file can use the information that I gave then it should be relatively simple to extract the version number and compare with the previous version and only generate a new packagebuild file if the version changes - that could even be automated and executed say daily? I was only trying to see if there was a way to both reduce the maintainer workload and also make the update perhaps more timely... I hope that is potentially useful?

Det commented on 2012-03-05 22:03 (UTC)

I meant that t3ddy mentioned some time ago to be using some script to generate the PKGBUILD for him. And the problem with auto-fetching PKGBUILD is that yes, it's better when the package would otherwise be out-of-date and it also eases up the maintainer's task but the whole point of _having_ a packaging system with linux is that the system/packages update automatically. With semi-automatically updating packages you'd first have to find out whether there even was a new version and then manually reinstall the package. Who wants that? E: also, I mistook you for ruario. That's why I figured you knew about the script thing.

mike.cloaked commented on 2012-03-05 21:43 (UTC)

If there is a way to utilise the downloaded file that I suggested, and extract the information required within the existing but amended PKGBUILD file then it would make life easier for the users? I don't know if it is possible but at least there is a chance to make the whole thing automatic rather than needing manual intervention by the maintainer each time chrome is updated? Of course I could just be too naive to know there is no simple way to make this work and would be happy to be informed whether or not my suggestion is possible or not.

mike.cloaked commented on 2012-03-05 20:40 (UTC)

Which script does t3ddy use? Why is it a better solution than having a self-contained script so that new amended copies are not necessary mostly? What is the problem with curl/awk anyway?

Det commented on 2012-03-05 18:43 (UTC)

If you're talking about replacing the "pkgver" line with some curl/awk odds and ends, you can forget it. If you're just talking about the script t3ddy is using, then that's a very nice solution.

mike.cloaked commented on 2012-03-05 18:26 (UTC)

I was wondering today if it might be possible to get at the version number and checksum information direct from the google repo, and by trial and error I have found today that it is possible to get at this information. By typing the command as in the following you can see that the right information is extracted: [mike@home1 chrome-test]$ curl -s |gzip -d| awk -F\" '/<name>google-chrome-stable<\/name><arch>/ { print $6 ; print $8 ; print $13}' 17.0.963.65 124586 >2d0a251fb70396a2943c85d58aa859e1392fd101</checksum><summary>Google Chrome</summary><description>The web browser from Google So in principle the "ver" and "rel" numbers are easy to extract and in addition the sha1sum can be extracted from the last line as well - and this could allow a single aur PGGBUILD file to be written that would extract all the required information without having to download a new tarball each time google updates chrome (unless they changed the format of the "primary.xml.gz file that contains the data. I wonder if a suitable PGKBUILD file could be constructed based on the above information? I am not that expert at awk commands and it might be quicker if someone really familiar with awk could do that? If so it would make for a really nice update to this AUR package, since the current file needs to be manually updated with the new version numbers and checksum every time google updates chrome? The checksum would need to be changed from md5 to sha1 but that is presumably not a huge change? A very similar scheme would presumably also work for beta and dev versions? Either way it is easy enough to download the "primary.xm.gz" file to tinker with the rest of the necessary scripting. I hope this might help? Thanks

commented on 2012-03-05 14:20 (UTC)

fixed PKGBUILD (new version):

dejavu commented on 2012-02-24 19:27 (UTC)

Ok, thanks for the answers. First I meant "chrome://plugins" ;-) I was interested in the built-in flash of chrome because of the advertised future of flash in linux. I am using 64 bit and so therefore it's not included. Thanks anyway ...

t3ddy commented on 2012-02-24 07:49 (UTC)

Chrome's flash plugin is available only for 32 bit's package. You've got to install flashplugin if you're on 64 bit.

commented on 2012-02-24 07:18 (UTC)

@t3ddy I not have flash installed and not have the plugin why the plugin native from g-chrome not is installed??

t3ddy commented on 2012-02-23 20:04 (UTC)

Do you mean in "about:plugins"? I've got it. Are you on 32 or 64 bit?

dejavu commented on 2012-02-23 19:46 (UTC)

Thanks for the package! But I can't see the flashplugin in "about:chrome" included.

mike.cloaked commented on 2012-02-16 14:13 (UTC)

That was quick - and thanks for updating.

mike.cloaked commented on 2012-02-16 11:08 (UTC)

Any chance this package will be updated to version 17.0.963.56-121963 which was released in the last day or so? Thanks.

Det commented on 2012-02-05 22:33 (UTC)

Of course it doesn't. That's repackaging not building.

Vrtak-CZ commented on 2012-02-05 22:29 (UTC)

/usr/bin/google-chrome: error while loading shared libraries: cannot open shared object file: No such file or directory after rebuild does not help :-(

commented on 2012-01-24 08:50 (UTC)

Build error.

commented on 2012-01-24 08:50 (UTC)

Build error.

t3ddy commented on 2012-01-14 09:29 (UTC)

Added man page. Now, I'll wait for a new release from Google to see what I've broken :)

Det commented on 2012-01-13 22:32 (UTC)

Would you?

t3ddy commented on 2012-01-13 22:15 (UTC)

Finally I've found the time to look at this. I've made script and template for rpm. There's one thing really strange: "chrome-sandbox" has wrong permissions, so I've to fix them in pkgbuild. Another thing: I've seen rpm's got also man page for chrome, shall I add it?

Det commented on 2012-01-13 15:26 (UTC)

@ruario, yeah, when it's extracting the sources. Just like it'd do with a regular tarball.

ruario commented on 2012-01-13 13:47 (UTC)

@t3ddy: actually it seems makepkg now opens rpms (using bsdtar) for you! ;)

ruario commented on 2012-01-09 13:00 (UTC)

@t3ddy: Some other thoughts and/or possible improvements: Rather than using a fixed set icons and duplicating each of them in /usr you could symlink all available icons, e.g. # Symlinking icons to /usr/share/icons/hicolor/ for png in "$pkgdir/opt/google/chrome/product_logo_"*.png; do pngsize="${png##*/product_logo_}" mkdir -p "$pkgdir/usr/share/icons/hicolor/"${pngsize%.png}x${pngsize%.png}/apps (cd "$pkgdir/usr/share/icons/hicolor/"usr/share/icons/hicolor/${pngsize%.png}x${pngsize%.png}/apps/; ln -s ../../../../../../opt/google/chrome/product_logo_${pngsize} google-chrome.png) done You could also consider using the rpm as a source rather than the deb. The rpm is more generic, unlike deb packages which has several debianisms. For example "usr/share/menu/" (used only on debian systems) and the compressed changelog.gz (containing only a link and is really only provided to satisfy lintian, debian's package quality checking script). It is very easy to open an rpm on Arch. You do not need rpm2cpio as all Arch installs include bsdtar, which is provided by libarchive on which Arch's own pkgtools depend. In fact makepkg uses bsdtar itself to open tar files. bsdtar can open rpms directly without requiring any rpm components (e.g. bsdtar xf google-chrome-stable_current_amd64.rpm) If you need to read rpm packaging meta-data locally and don't have rpm installed you can just use "strings google-chrome-stable_current_x86_64.rpm | less". For you personal script that you use to update the PKGBUILD, this would tell you the latest version and build number: _verbld=$(wget -qO-$(uname -m | sed 's/i.86/i386/')/repodata/other.xml.gz | zcat | awk -F\" '/-stable/ { print $10"-"$12 }') Thus making the direct download urls:${_verbld}.i386.rpm${_verbld}.x86_64.rpm Whether the deb or rpm is used the file "/etc/cron.daily/google-chrome" should be removed from Arch before packaging as it is a cron job to use Google's own rpm or deb repositories and hence is pointless on Arch. No rush and feel free to ignore them all. Just some ideas. ;)

t3ddy commented on 2012-01-09 12:42 (UTC)

ruario is right, I've got a script that acts on a template to make the pkgbuild. Now I don't have time to look at your suggestion, but I'll do asap. Thanks :)

ruario commented on 2012-01-09 12:16 (UTC)

> you can remove the unnecessary ones by just using the red/white cross buttons there. Done > he's not using any scripts. I beg to differ. It seems like he is his own script locally (not shared here) to create/update the PKGBUILD when updates to Chrome happen. Quotes from t3ddy, "I could change the pkgbuild if there's a way to predict the build number automatically, so I can change the script to read the version and modify the pkgbuild accordingly." "The problem is that chrome gets updated very frequently, so I'm using a script to do the job. Making your change will make the script useless and increase my work." At a guess, when he is notified (or becomes aware of an update by some other means), he runs a script which downloads the latest version for both architectures, checks the md5sums and updates the PKGBUILD. As I said, this makes perfect sense to me. I do the same with my own frequently updating PKGBUILD. His complaint previously about changing the URL to point to a specific build that this would make his life harder because it "will make the script useless".

Det commented on 2012-01-09 11:23 (UTC)

Well first off you can remove the unnecessary ones by just using the red/white cross buttons there. Also he's not using any scripts. The URL he's using doesn't change - just the file that gets downloaded.

ruario commented on 2012-01-09 09:25 (UTC)

@DetYes, apologies. I thought of each improvement a bit at a time. I wasn't international spamming. Anyway I don't think the idea would be to put it in the PKGBUILD itself. t3ddy said, "The problem is that chrome gets updated very frequently, so I'm using a script to do the job." I read that as him having some meta-script that creates/updates the PKGBUILD for him. This make sense as I do the same thing for opera-next (opera development builds). So presumably he would add the logic to his meta script, not the PKGBUILD itself.

Det commented on 2012-01-09 08:41 (UTC)

1) And that couldn't all fit in one post? 2) Also that's gonna look so great in the Sources section there.

ruario commented on 2012-01-09 08:06 (UTC)

@t3ddy: I suspect this is what you want your meta script to do: echo$(curl -s | zcat | awk -F "[ -]" '/Package: google-chrome-stable/ { getline ; print $2"-"$3 }')_$(uname -m | sed 's/i.86/i386/;s/x86_64/amd64/').deb It will print back whatever is the link to the current stable version.

ksmanis commented on 2012-01-07 01:11 (UTC)

PLEASE add kdebase-kdialog as an optdepend! Took me ages to figure out which package was needed.

commented on 2012-01-03 16:06 (UTC)

Important for all KDE-Users: Chrome uses from version 16 on the default KDE-dialogues. Therefore the packages kdebase-kdialog has to be installed. Otherwise Chrome won't show any dialog-box under KDE.

t3ddy commented on 2011-11-16 20:02 (UTC)

I could change the pkgbuild if there's a way to predict the build number automatically, so I can change the sript to read the version and modify the pkgbuild accordingly.

Det commented on 2011-11-16 18:55 (UTC)

T3ddy is already updating this thing every time someone flags it. He's just talking about not needing to change the build number when doing so.

commented on 2011-11-16 18:11 (UTC)

I agree with tarpman, as that is the way PKGBUILD's are supposed to work. Even beyond that, I'm sure someone can flag it out of date if you miss an update or forget. As a maintainer, I find that's the much more easier and lazier way of doing it: only update when someone pesters me lol

t3ddy commented on 2011-11-12 20:39 (UTC)

I like your idea. The problem is that chrome gets updated very frequently, so I'm using a script to do the job. Making your change will make the script useless and increase my work. For now I won't make changes, anyway thanks for your suggestion :)

commented on 2011-11-12 17:51 (UTC)

Instead of taking the current version from the direct DL link and having your md5sum fail when it changes, you could point the PKGBUILD at the versioned deb from the pool instead. e.g. When they clean out old versions that will start to 404 (and also directory listings aren't enabled so you have to know the exact filename you want to download) but I prefer that to an md5sum failure.

commented on 2011-11-10 17:28 (UTC)

Add dependencies xdg-utils becouse not open downloads folder.

Det commented on 2011-10-27 10:41 (UTC)

Or you can just check the version number to see whether it matches with the latest one: I'd prefer building with 'makepkg -si --skipinteg' anyway, since 1) the chances of getting a corrupt download from Google servers in practice are nonexistant and 2) you'd be using the md5sum of your own corrupt download anyway.

commented on 2011-10-27 09:24 (UTC)

If the md5sum fails, it is probably because the *.deb package was updated. If you are certain your *.deb download is OK, you can manually replace the appropriate md5sum in the PKGBUILD or use the command posted below by nDray. Google-chrome is updated very frequently, so md5sums are difficult for the package maintainer to keep current.

Det commented on 2011-10-21 16:17 (UTC)

@massol, that's not a fix - it does the same thing. Clear out your package cache first.

commented on 2011-10-21 15:22 (UTC)

if md5sum failed, then try yaourt -Syu

unkier commented on 2011-10-07 04:09 (UTC)

yep, md5sum failed

thesame commented on 2011-09-20 19:24 (UTC)

md5sum failed

nDray commented on 2011-09-05 09:00 (UTC)

makepkg -g >> PKGBUILD every new release

kardia commented on 2011-09-04 23:45 (UTC)

md5sum failed

commented on 2011-08-26 02:36 (UTC)

md5sum failed

t3ddy commented on 2011-08-25 15:31 (UTC)

It still does: mv -v "$pkgdir/opt/google/chrome/google-chrome.desktop" "$pkgdir/usr/share/applications"

canuckkat commented on 2011-08-25 14:15 (UTC)

Is it just me or did this package used to make a menu item?

nDray commented on 2011-08-09 22:37 (UTC)

It's wrong because you need to check the new release numbers, it always fetches the latest. Before "makepkg" run "makepkg -g >> PKGBUILD"

HNKNTA commented on 2011-08-09 22:30 (UTC)

md5sum at least for x86_64 is wrong

commented on 2011-08-02 16:57 (UTC)

The latest stable version is now 13.0.782.107 ;)

duca commented on 2011-06-15 21:44 (UTC)

Just to add the info as well, the current as of today md5sum for i386 version is 41b92b2f35f12cf2e31998a22e9dba7a

Commander commented on 2011-06-15 00:08 (UTC)

New md5sum for 12.0.742.100 409c1d33d38e09bf295f8e57e18d210b google-chrome-stable_current_amd64.deb

commented on 2011-05-04 12:09 (UTC)

As pciutils came to [core], now you can remove libpci dependency.

Det commented on 2011-04-07 13:33 (UTC)

@unknown, still, libpci is yet only provided by the [testing] pciutils. E: and what I'm talking about is that the package doesn't say that it provides it.

unknown commented on 2011-04-07 09:47 (UTC)

libxtst is missing from the dependencies. libpci is already provided by pciutils.

Det commented on 2011-04-06 08:28 (UTC)

@serraemeira, not until the new [testing]'s pciutils hits [core]. Until then you can just remove the indication to 'libpci' in the dependencies line. Btw. did anybody notice that the bug report t3ddy made about this was closed for the wrong reason?: Closed by Andrea Scarpino (BaSh) Friday, 25 March 2011, 21:05 GMT-4 Reason for closing: Not a bug Additional comments about closing: the libpci package should conflicts with pciutils Because they conflicting with each other is not the issue. The issue is that pciutils doesn't provide libpci that it conflicts with (actually it doesn't even conflict with libpci - libpci conflicts with pciutils).

commented on 2011-04-06 03:02 (UTC)

Please, libpci should be dropped. Thanks!

commented on 2011-03-26 01:16 (UTC)

When pciutils now in testing enters core the libpci dependency should be dropped:

commented on 2011-03-08 17:10 (UTC)

The fast update, thanks.

commented on 2011-02-03 19:10 (UTC)

New checksum for x86 is 41e6f6bb93d6bf98ef06941f0ee10194 and version is 9.0.597.84.

Det commented on 2010-12-17 13:33 (UTC)

Right.. didn't I just say so?

commented on 2010-12-17 09:14 (UTC)

last version is 8.0.552.224

Det commented on 2010-12-15 09:16 (UTC)

This package is auto-updating so in cases like these it's enough if you just redownload the whole package. You can additionally change the pkgver to "8.0.552.224" (the current one) if you wish, but that's not gonna change how the package functions. It's just a note for pacman.

commented on 2010-12-08 03:32 (UTC)

Hi. May you please update this PKGBUILD? Thanks in advance...

master commented on 2010-11-01 19:36 (UTC)

Package fixing Google Chrome's crash on opening jpgs:

Det commented on 2010-10-31 13:09 (UTC)

Heh, interestingly the same mistake of using 7.0.517.43 instead of .41 was done in [extra] too (3 days ago). Looks like it was flagged as "out-of-date" shortly after. As for your case, actberw, first off, it's not a crime to use spaces after commas. Seccondly you don't need to worry about installing libjpeg6 despite you already got libjpeg(8). They are installed side-by-side and whatever application uses either libjpeg only uses _that_ one.

commented on 2010-10-31 12:44 (UTC)

it need libjpeg6,but by default,my arch install libjpeg8,what should I do?

Det commented on 2010-10-25 16:45 (UTC)

Yeah, I already notified x-demon about it. He accidentally checked in "" that the latest stable "Google Chrome" would be 7.0.517.43, while that's actually the version number of the latest stable "Google Chrome _Frame_". Chrome is still at 7.0.517.41. I'd also like to see him updating "google-chrome-beta" _even though_ the PKGBUILD is "auto-updating" :s.

commented on 2010-10-25 12:09 (UTC)

The version number should be adjusted. After all, the latest stable version is 7.0.517.41. But it makes no difference for the installation.

Det commented on 2010-10-21 17:03 (UTC)

A lot of crashes should've been fixed with 7.0.517.41 - did you update the package? You can do this by just redownloading it (the package is auto-updating).

blackdeagle commented on 2010-10-20 18:21 (UTC)

got the same error like nDray. And the crashing issue is VERY annoying, hope it'll get fixed soon =)

nDray commented on 2010-10-19 19:15 (UTC)

==> Preparing install `opt/google' -> `/home/ndray/tmp/dev/google-chrome/pkg/opt/google' -> Done preparing! ==> Patching launcher patching file google-chrome Hunk #1 FAILED at 30. 1 out of 1 hunk FAILED -- saving rejects to file google-chrome.rej Aborting...

Det commented on 2010-09-22 14:56 (UTC)

Because the (.)debs this package uses are built for the buggy, Ubuntu favored libjpeg6. There would be no sense in having a old dependency that isn't required by the package - without it there would be no viewing of jpgs with this package at all. We are a bleeding edge distribution and using libjpeg 8. That's why chromium-browser-svn won't crash with (.)jpgs.

commented on 2010-09-19 19:05 (UTC)

Then what's the reason of crasing while attaching *.jpg files?

Det commented on 2010-09-19 17:26 (UTC)

@elfaimer, libjpeg6 _doesn't_ conflict with [extra]'s libjpeg.

commented on 2010-09-19 00:51 (UTC)

Yes, I mean .jpg files. I don't want to compile anything else, I use Firefox for that job. It is just annoying.

commented on 2010-09-18 22:23 (UTC)

samurai, do you mean .jpg files? Apparently, libjpeg6 as chrome dependency conflicts with libjpeg 8.0.2, installed by default. Compiling chromium-browser-svn emerged as solution for me.

commented on 2010-09-18 14:02 (UTC)

By the way, is there a solution for attached images crash?

Det commented on 2010-09-18 06:32 (UTC)

[extra]'s chromium also depends on bzip and libevent.

commented on 2010-09-17 23:49 (UTC)

I installed xdg-utils package and now I can make Chrome default browser :) Maybe, you should add it as dependency.

commented on 2010-09-17 17:52 (UTC)

I can't make Google Chrome my default browser. Any idea?

commented on 2010-09-10 08:28 (UTC)

Please update browser thanks.

trusktr commented on 2010-08-25 13:59 (UTC)

@WickedOne hmmm, that's strange, mine had everything!

trusktr commented on 2010-08-25 13:58 (UTC)

This package should be moved to the official repos to be along side firefox, etc!

commented on 2010-06-15 02:54 (UTC)

When i installed this...none of the icons are installed to the system. Also it does not install any menu entries in the gnome menu.

commented on 2010-06-11 15:42 (UTC)

Is libjpeg6 and libpng12 still needed? Chromium (same version) compiles against the newer versions...

print commented on 2010-05-26 01:40 (UTC)

Here's the official announcement stating that chrome for linux is stable:

commented on 2010-05-25 20:28 (UTC)

Not sure if I followed the standard protocol, but Google released a stable version of Chrome and I thought there should be a package specifically for it. I just copied the beta package and changed a few lines.