Package Details: google-chrome 52.0.2743.116-1

Git Clone URL: https://aur.archlinux.org/google-chrome.git (read-only)
Package Base: google-chrome
Description: An attempt at creating a safer, faster, and more stable browser (Stable Channel)
Upstream URL: https://www.google.com/chrome
Licenses: custom:chrome
Provides: google-chrome, pepper-flash
Submitter: None
Maintainer: Det
Last Packager: Det
Votes: 1338
Popularity: 36.303352
First Submitted: 2010-05-25 20:25
Last Updated: 2016-08-04 13:25

Required by (18)

Sources (3)

Pinned Comments

Det commented on 2016-03-09 05:01

✔ NOTE: If the md5sums don't match, don't post the makepkg output. Simply flag the package.

You can check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Stable%20updates, or use:
$ curl -s https://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }'

Det commented on 2016-03-02 20:53

✔ OBS: 32-bit support for Google Chrome Stable Channel has been dropped as well. Either upgrade to 64-bit, or stick to [extra]/chromium, which will continue to be updated:

- http://linux.slashdot.org/story/15/12/01/1524259/google-to-drop-chrome-support-for-32-bit-linux

Latest Comments

Det commented on 2016-08-04 14:15

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

> 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pacaur -Sau run so I decided to look into here what's going on and post my experience since the package was marked out of date since yesterday and nothing interesting happened here.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

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

> How so?

http://pastie.org/private/uxa5jb2ossm8dizsg4zs0a

> ✔ 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

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 Your package google-chrome [1] has been flagged out-of-date by hgaronfolo [2]:

> Out of date.

How so?

$ curl -s https://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | 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

Dooone.

LuigiR0jas commented on 2016-07-31 17:20

I had the same font issue as @rthundr, and solved it the same way.
gnome-themes-standard should really become dependencies for this package

rthundr commented on 2016-07-25 18:41

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

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

@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

Strangely one day after I've updated again and now it's running fine.

swanson commented on 2016-07-21 15:39

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

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

I am noticing that youtube videos are playing but I am getting a blank screen.

Unia commented on 2016-05-15 19:32

Since pacman introduced hooks, running update-desktop-database from the .install file is no longer necessary.

johncgage commented on 2016-05-12 01:30

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

4.5.3 but I was having the same issue on 4.5.2

johncgage commented on 2016-05-12 00:39

What kernel version are you running?

keiren commented on 2016-05-12 00:28

I'm getting this when I try to run "google-chrome-stable"
[1:1:0512/102613:ERROR:nacl_fork_delegate_linux.cc(315)] Bad NaCl helper startup ack (0 bytes)

Det commented on 2016-04-29 06:59

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

Will it be fixed in the near future?
Need the stable version of Google Chrome.

Det commented on 2016-04-29 06:01

Because it "is out-of-date".

Click the red link for the message.

Svante commented on 2016-04-29 05:32

Why is google-chrome "Flagged out-of-date"?

darkxsun commented on 2016-04-18 21:18

@mol3earth This should help you get started with the Arch User Repository:

https://wiki.archlinux.org/index.php/Arch_User_Repository#Getting_started

mol3earth commented on 2016-04-18 20:02

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

@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

Could you please save the eula_text.html file into the package?
In China, www.google.com is blocked (while dl.google.com 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

Nice tip. It's got them md5sums too.

z3ntu commented on 2016-04-07 19:38

FYI:
$ curl https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-amd64/Packages | 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

Upstream problem.

morantron commented on 2016-04-07 17:57

This link is broken and the install no longer works https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb

bcroq commented on 2016-03-29 10:18

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

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

Please put all of it in Pastebin, and you should also use crbug.com.

resetko commented on 2016-03-10 21:16

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 http://pastebin.com/5VcgEjuh)

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

✔ NOTE: If the md5sums don't match, don't post the makepkg output. Simply flag the package.

You can check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Stable%20updates, or use:
$ curl -s https://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }'

Det commented on 2016-03-03 08:22

Kill it first.

justinnichols commented on 2016-03-03 00:40

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

@xtrm0, "unmatching md5".

No it isn't.

Det commented on 2016-03-02 20:53

✔ OBS: 32-bit support for Google Chrome Stable Channel has been dropped as well. Either upgrade to 64-bit, or stick to [extra]/chromium, which will continue to be updated:

- http://linux.slashdot.org/story/15/12/01/1524259/google-to-drop-chrome-support-for-32-bit-linux

maxspice commented on 2016-03-02 20:42

@Det, sorry. That was stupid of me.

Det commented on 2016-03-02 20:40

But please don't flag it _after_ it's updated.

Det commented on 2016-03-02 20:36

Read the description, please.

maxspice commented on 2016-03-02 20:35

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

Oбновляются. Воспользуйтесь следующий раз Google Translate: https://goo.gl/kbgGPG

Updated. Please Google Translate next time: https://goo.gl/kbgGPG

cyberlinux commented on 2016-02-09 22:33

Не совпадают md5sums

Спасибо с Уважением Павел

cyberlinux commented on 2016-02-09 22:29

[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

Hello.

The md5sums don't match in:
google-chrome-stable_48.0.2564.103_amd64.deb

Det commented on 2016-02-05 13:00

Easiest way to test is always with a clean profile. Rename ~/.cache/google-chrome/ and ~/.config/google-chrome/.

ugjka commented on 2016-02-05 12:54

Anyone experiencing a total freeze when opening the Downloads page while downloading something?

Det commented on 2016-02-04 09:27

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

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

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

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

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

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

@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

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

The license prevents it: https://www.google.com/chrome/browser/privacy/eula_text.html

tago commented on 2015-12-24 16:31

Why is not on official rep?

Det commented on 2015-12-12 11:14

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

This package shows up on the list of packages that need to be rebuit after the recent switch to C++ dual ABI.
https://lists.archlinux.org/pipermail/arch-dev-public/2015-December/027597.html
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

So it looks like the problem I'm having is a known issue over at the Chromium project.

https://code.google.com/p/chromium/issues/detail?id=525751

The problem is some older CPU don't have the LAHF instruction.

schultzter commented on 2015-11-04 20:44

[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

Rename ~/.config/google-chrome and ~/.cache/google-chrome, or have a look at the terminal output.

ricefan123 commented on 2015-10-31 09:39

Unable to run the application of current version

petterk commented on 2015-10-13 19:19

NVIDIA: could not open the device file /dev/nvidia0 (Operation not permitted)

Det commented on 2015-10-12 08:05

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

without the "- " in front, it was meant to indicate that this line should be replaced.

Det commented on 2015-10-07 17:43

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

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

@jeffska, you're 2 days late.

jeffska commented on 2015-09-25 01:58

@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

@elemongw, why?

Det commented on 2015-09-09 23:29

Yepyepyep.

YepyepyepYepyepyepYepyepyepYepyepyep.

Yepyep.

m45t3r commented on 2015-09-09 12:24

@BrtH Yep, according to Docky wiki, the WMClass in the current .desktop file is wrong:
http://wiki.go-docky.com/index.php?title=How_to_Customize_Window_Matching

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

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

Thanks.

Det commented on 2015-09-04 19:00

I mean "no" to the "no". Link the thread next time: https://bbs.archlinux.org/viewtopic.php?id=201911

Didn't bump the pkgrel. It's quicker to add the quotes yourself.

ugjka commented on 2015-09-04 18:40

https://bbs.archlinux.org/viewtopic.php?id=201911

Det commented on 2015-09-04 18:28

No.

Det commented on 2015-09-04 18:28

No.

ninian commented on 2015-09-03 22:25

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?

SupKurtJ commented on 2015-09-02 11:50

it's ok now with google-chrome-stable_45.0.2454.85_amd64.deb
thanks

SupKurtJ commented on 2015-09-01 20:31

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

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

@Det, You have 'pacman' in 'makedepends'. pacman is included to base-devel group.
https://wiki.archlinux.org/index.php/PKGBUILD#makedepends
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

What..?

lybin commented on 2015-08-28 04:49

https://wiki.archlinux.org/index.php/PKGBUILD#makedepends
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

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

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

It's .157.

T1000 commented on 2015-08-20 20:32

Version: 44.0.2403.155-1
md5sums_x86_64 checksum needs to be changed to 24097e63c43976e9f9304edd5bcd4ac0

Det commented on 2015-08-14 23:28

@hewr1993, why..?

Det commented on 2015-08-12 13:11

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

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::http://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-stable/google-chrome-${_channel}_${pkgver}-1_i386.deb")
source_x86_64=("google-chrome-${_channel}_${pkgver}_amd64.deb::http://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-stable/google-chrome-${_channel}_${pkgver}-1_amd64.deb")

Det commented on 2015-08-11 18:09

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

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

lol

j3sk0la commented on 2015-08-11 17:58

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

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

I just tried launching .107 a couple of times, and it works just fine so far.

Det commented on 2015-07-24 18:36

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-*): http://mirror.pcbeta.com/google/chrome/deb/pool/main/g/google-chrome-stable/google-chrome-stable_43.0.2357.134-1_amd64.deb

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

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: http://sprunge.us/cePj.

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

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

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

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

Absolutely no need, when you can just do:

$ makepkg -si --skipinteg

toppy commented on 2015-07-14 21:29

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

@tmoore, 1) I know, this was flagged (see top of PKGBUILD), and 2) see: https://wiki.archlinux.org/index.php/PKGBUILD#Integrity

tmoore commented on 2015-07-14 20:05

New md5sum for google-chrome-stable_43.0.2357.132_amd64.deb is b26558f6be7834d36a58e70348082203

Also, probably should switch pkgbuild to use sha256sum?

adventurer commented on 2015-07-14 14:06

Updates to v. 43.0.2357.134

darkxsun commented on 2015-06-22 16:02

Flagging out-of-date is good enough...

CurtisLeeBolin commented on 2015-06-22 15:45

$ 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

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

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!

darkxsun commented on 2015-05-27 16:19

@costis: Kill all chrome processes. You've got one left lying around in the background.

Det commented on 2015-05-27 16:19

Make sure you kill the processes first.

costis commented on 2015-05-27 16:18

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

@sli908, but that's awesome that they have a bug tracker for that kinda thing ^^: https://code.google.com/p/chromium/issues/detail?id=478714

sli908 commented on 2015-05-26 02:16

Ok the issue was fixed in the latest update (at least in MATE).

@Det We are trying to see what platforms are affected by the issue. If you have nothing constructive to contribute, please do not comment. Especially in a juvenile manner such as the way you did.

Det commented on 2015-05-24 05:22

ohmygod

nagar commented on 2015-05-23 23:16

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

Of COURSE it is, stop SAYING it!

marmotz commented on 2015-05-23 14:12

@sli908 @isra17 same issue here with MATE

sli908 commented on 2015-05-22 01:59

@isra17 I checked and Chromium seems to be having the same issue.

isra17 commented on 2015-05-21 23:26

@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

@sandnon thanks that works for me

oryxfea commented on 2015-05-21 07:19

@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

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

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?

petterk commented on 2015-05-20 15:54

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

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

@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

@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...

sinatosk commented on 2015-05-20 14:06

@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 srt to very small or small even when I set to very small

petterk commented on 2015-05-20 12:00

Anyone else have weird looking fonts after the last update?

Det commented on 2015-05-20 06:43

Flag it.

samehkamaleldin commented on 2015-05-20 06:42

==> 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

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

also here: http://ftp.ubuntu-tw.org/mirror/deepin/pool/universe/g/google-chrome-stable/

definitely an issue with upstream.

Det commented on 2015-05-19 19:18

Yeah, they're all available in: http://mirror.pcbeta.com/google/chrome/deb/pool/main/g/google-chrome-stable/

See, if that helps.

madscience commented on 2015-05-19 19:15

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

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

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

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

Remove the "Version" line from the .desktop (/usr/share/applications/google-chrome.desktop).

madscience commented on 2015-05-19 18:25

I didn't see any suggestion in the link to the git commit...?

Det commented on 2015-05-19 18:21

No, but did you actually do what was suggested in the link?

madscience commented on 2015-05-19 18:18

ran both of those as root, logged out and back in... same issue.

Det commented on 2015-05-19 18:11

Does that help?: https://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/chromium&id=5ba771828dd14d62b6794006c0bb6d3cec3ac0e6

Run "update-desktop-database" and maube "xdg-icon-resource forceupdate --theme hicolor" as root afterwards.

madscience commented on 2015-05-19 18:09

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.

kamelie1706 commented on 2015-05-17 09:43

Forget about it ... was still having one instance running in the background!

kamelie1706 commented on 2015-05-17 09:39

Hi,

I installed the latest version
google-chrome-42.0.2311.152-1-x86_64.pkg.tar.xz

But when I checked from chrome I get
Version 38.0.2125.104 (64-bit)

If I list foreign package I see only one google chrome package
google-chrome 42.0.2311.152-1


What am I missing here?

Det commented on 2015-05-12 18:27

Wow. 1000 votes: http://i.imgur.com/yTqUGY2.jpg

Det commented on 2015-05-12 18:27

Wow. 1000 votes: http://i.imgur.com/yTqUGY2.jpg

Krieghof commented on 2015-04-16 18:00

@Det Thank you very much. My / partition was full. Now it works flawless.

Det commented on 2015-04-16 17:14

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: http://community.skype.com/t5/Linux-archive/Skype-does-not-start-Bus-Error-core-dumped/td-p/1423196

Krieghof commented on 2015-04-16 16:42

Can't start chrome after the latest update.
The Terminal says "Bus error (core dumped)"

http://i.imgur.com/MHeB8xD.png

oryband commented on 2015-04-09 11:53

@sugatang.itlog - if you have scroll and mouse pointer issues, check the following chromium issue: https://code.google.com/p/chromium/issues/detail?id=465660

They're talking about a temporary fix, take a look.

ekkelett commented on 2015-04-08 05:47

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

-> Downloading google-chrome-stable_41.0.2272.118_amd64.deb...
Initializing download: https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb
HTTP/1.0 404 Not Found

Something is wrong.

sugatang.itlog commented on 2015-04-05 06:08

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

new md5sums:
ef61d7c2b66ba26b1f297f82e7e1a52b google-chrome-stable_current_amd64.deb
3d9b5f96dbdeaca079b39d41135b662b google-chrome-stable_current_i386.deb

fichtner commented on 2015-04-01 18:56

md5sums currently out of date. Was able to install with modifying PKGBUILD

zman0900 commented on 2015-03-20 04:29

└──╼ curl -s https://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | 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

Still .89:

0 ✓ det_000@Awesomeness-PC /cygdrive/c/Users/det_000/Desktop
$ curl -s https://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | 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

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.

Det commented on 2015-03-15 01:17

What types of files are you trying to open that don't show Chrome in the menu?

dejavu commented on 2015-03-12 21:59

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

Thank you Det!

JetJaguarXP commented on 2015-03-11 00:06

I don't see any buttons labeled "Flag package out-of-date". There is a link with that name on this page. If that is what you are referring to, what does that do?

Det commented on 2015-03-11 00:03

Stop, please.

Det commented on 2015-03-11 00:02

Well, you don't have to flag it _anymore_, I already updated it.

Det commented on 2015-03-10 23:27

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

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

Thanks a lot.

cellisten commented on 2015-03-03 19:58

New md5sum and also changed from data.tar.lzma to data.tar.xz

Det commented on 2015-03-01 16:00

@shibumi, why?

brittyazel commented on 2015-02-28 09:35

And I disable all extensions :-/

brittyazel commented on 2015-02-28 09:35

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

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

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

You know, "menu → Settings → Manage search engines...", which are saved in the 'Web Data' database.

brittyazel commented on 2015-02-27 23:18

What do you mean by one of my search engines?

Det commented on 2015-02-27 23:08

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

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

Shouldn't you clear your cache then?

brittyazel commented on 2015-02-27 23:00

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

AJSlye, reason?

WorMzy commented on 2015-02-19 16:50

40.0.2214.115 just out: http://googlechromereleases.blogspot.co.uk/

Det commented on 2015-02-12 07:13

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/

Det commented on 2015-02-12 07:13

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

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

@mgb, that's for Chrome OS: http://googlechromereleases.blogspot.com

Det commented on 2015-02-07 17:10

This is true.

weirddan455 commented on 2015-02-07 17:01

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

@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..."?

Det commented on 2015-02-07 16:32

@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 walk 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

/opt/google/chrome/chrome: error while loading shared libraries: libudev.so.0: cannot open shared object file: No such file or directory

weirddan455 commented on 2015-02-07 14:28

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

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: https://wiki.archlinux.org/index.php/PKGBUILD#Integrity_variables

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] = https://wiki.archlinux.org/index.php?title=PKGBUILD&diff=359592&oldid=359584
[2] = https://wiki.archlinux.org/index.php/Talk:PKGBUILD#pkgdir

tormund commented on 2015-02-06 18:10

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

You need to press "Flag package out-of-date" from the right :).

Xiaoming94 commented on 2015-02-05 17:23

plz update this package

Det commented on 2015-02-05 10:11

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 (https://en.wikipedia.org/wiki/MD5), 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: https://wiki.archlinux.org/index.php/PKGBUILD

vikstrous commented on 2015-02-05 06:16

Please use something better than md5, like sha256.

Det commented on 2015-02-04 13:50

This wasn't really all that hard to find: https://code.google.com/p/chromium/issues/detail?id=430910

It's got to do with libgtk indeed.

gauravpadia commented on 2015-02-04 13:44

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 libgtk-x11-2.0.so.0.2400.25[7ffd76e31000+436000]
[12221.628406] chrome[5593]: segfault at 30 ip 00007f73be130186 sp 00007fff75c3e9a0 error 4 in libgtk-x11-2.0.so.0.2400.25[7f73bde71000+436000]

Det commented on 2015-02-04 00:49

You can use:

$ curl -s https://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | 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

checksum seems to be incorrect, for me (64 bit) the correct md5sum is
361aea21e6e3acaa805af17edb4ffe49

Det commented on 2015-01-29 22:02

Fixed.

tigrang commented on 2015-01-29 21:49

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

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

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

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

@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

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 libgtk-x11-2.0.so.0.2400.25[7f8fd8ee5000+435000]
[172071.870967] chrome[3162]: segfault at 30 ip 00007f077a445956 sp 00007fff00e16fe0 error 4 in libgtk-x11-2.0.so.0.2400.25[7f077a187000+435000]
[243127.630583] chrome[28297]: segfault at 30 ip 00007f088b43f956 sp 00007fff59e99f90 error 4 in libgtk-x11-2.0.so.0.2400.25[7f088b181000+435000]
[607124.690762] chrome[24993]: segfault at 30 ip 00007f6ce0706956 sp 00007fffd6963c30 error 4 in libgtk-x11-2.0.so.0.2400.25[7f6ce0448000+435000]

Det commented on 2015-01-27 13:04

@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 and...it doesn't have to be removed...: https://lists.archlinux.org/pipermail/aur-requests/2015-January/004440.html

LarcBP commented on 2015-01-27 12:06

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

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

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

Pacman 4.2.0 is apparently in [testing] for Manjaro: http://mirror.premi.st/manjaro/testing/core/x86_64/

You can get the previous (4.1.x compatible) PKGBUILD in: http://pkgbuild.com/git/aur-mirror.git/tree/google-chrome/PKGBUILD?id=670b3f5f6cc45f9a28652316ba17d02b60023cf3

annhill commented on 2015-01-16 09:24

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?

evaryont commented on 2015-01-13 21:13

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

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

Even with ttf-dejavu and ttf-liberation installed, I had Chinese characters show up as boxes. Tested on http://zh.wikipedia.org. I had to install ttf-droid to get them to display properly.

Perhaps another optional dependency?

rafaelff commented on 2014-12-10 15:18

@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

Nothing at all to do with this package.

grayich commented on 2014-12-10 09:52

wtf? sudo??

==> Leaving fakeroot environment.
==> Finished making: google-chrome 39.0.2171.95-1 (Wed Dec 10 11:51:21 EET 2014)
==> Exporting google-chrome to /var/cache/pacman/pkg/ directory
[sudo] password for gray:

Det commented on 2014-12-09 19:34

The servers still offer .71 for Linux: http://googlechromereleases.blogspot.com/2014/12/stable-channel-update.html

Det commented on 2014-11-28 17:45

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

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

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

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

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

I'm getting http error 404 on downloading the sources tarball.

Det commented on 2014-11-03 18:43

Apparently a pretty old bug. Couldn't reproduce with Beta or Dev channels: https://code.google.com/p/chromium/issues/detail?id=354571

CarstenG commented on 2014-11-03 18:28

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

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

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.
[17667:17735:1027/095325:ERROR:get_updates_processor.cc(240)] PostClientToServerMessage() failed during GetUpdates
[17667:17701:1027/095329:ERROR:channel.cc(316)] 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

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

Did you try https://www.google.com/search?q=ERROR%3Aget_updates_processor.cc+PostClientToServerMessage first?

Det commented on 2014-10-22 22:58

Did you try https://www.google.com/search?&q=ERROR%3Aget_updates_processor.cc(240)%5D+PostClientToServerMessage() first?

cyrilccros commented on 2014-10-22 22:56

Hi, google-chrome crashes whenever I load a page. This happens with chromium too, but Firefox is fine. I get an error in

...:ERROR:get_updates_processor.cc(240)] 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

It's now an optional dependency.

catalin.hritcu commented on 2014-10-18 13:27

Chrome doesn't display many PDFs correctly without ttf-liberation (https://code.google.com/p/chromium/issues/detail?id=369991). Shouldn't that be a package dependency?

Det commented on 2014-10-16 11:16

Please Google first: https://www.google.com/search?q=NVIDIA%3A+could+not+open+the+device+file+%2Fdev%2Fnvidia0+Operation+not+permitted

(E.g. the very first results).

Det commented on 2014-10-16 11:16

Please Google first: https://www.google.fi/search?q=NVIDIA%3A+could+not+open+the+device+file+%2Fdev%2Fnvidia0+(Operation+not+permitted)

(E.g. the very first results).

zlopi commented on 2014-10-16 08:40

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?

binarysage commented on 2014-10-15 04:09

Updated PKGBUILD :
https://gist.github.com/binarysage/8888d6492f9a194b9541

yodalee commented on 2014-10-15 02:43

The hash of google-chrome.deb should be
a2b61a623da973f12e1cb18674b9fb93
now

lviggiani commented on 2014-10-08 07:32

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

Same as with all the other OS's: https://code.google.com/p/chromium/wiki/ChromiumBrowserVsGoogleChrome

Barry commented on 2014-10-01 03:07

What's the difference between chrome and chromium on Arch?

Det commented on 2014-09-25 11:42

@corvinusz: "The stable channel has been updated to 37.0.2062.124 for _Windows_ and _Mac_."

- http://googlechromereleases.blogspot.com/2014/09/stable-channel-update_24.html

capoeira commented on 2014-09-21 13:02

Oh I see. now I just need to convince my bank to stop using it lol

Det commented on 2014-09-21 12:43

No it's not blocked, the whole support for the plugin API is gone.

capoeira commented on 2014-09-21 11:29

i understand that Java is blocked...the thing is, it's not even in chrome://plugins/

Det commented on 2014-09-20 20:09

It (NPAPI support) has been gone from Linux builds since v35: http://blog.chromium.org/2013/09/saying-goodbye-to-our-old-friend-npapi.html

capoeira commented on 2014-09-20 20:00

for the last 2 builds java doesn't apear anymore in about:plugins.......anyone?

cleanrock commented on 2014-09-12 12:43

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

"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

I really am unable to help with that.

passchaos commented on 2014-09-11 08:17

Help!!When I build chrome, I get "an error in package()"

phoenixlzx commented on 2014-09-11 02:48

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

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.

CharlesAtum commented on 2014-09-06 03:34

@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

Yeah. It was confirmed on Fedora also in your report. Also using Gnome Shell.

CharlesAtum commented on 2014-09-04 17:11

Anyone here also affected by this issue: http://crbug.com/381732 ?

wisecrick commented on 2014-09-02 16:08

This line
curl -Ls https://www.google.com/intl/en/chrome/browser/privacy/eula_text.html \
-o "$pkgdir"/usr/share/licenses/google-chrome/eula_text.html

hangs, as https://www.google.com/ is blocked

keiseele commented on 2014-08-31 19:08

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 19:07

keiseele commented on 2014-08-31 19:00

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

https://www.dropbox.com/s/tmg84j725o0hkh1/milocal2.png

wisecrick commented on 2014-08-31 12:50

Always hangs at
==> Starting package()...
-> Extracting the data.tar.lzma
-> Moving stuff in place

I don't know why...

Kuci commented on 2014-08-26 22:28

Did you forget to remove libgcrypt15 from dependencies ?

Det commented on 2014-08-13 12:36

Already updated.

niklas.heer commented on 2014-08-13 12:35

New md5sum: "b00bd4f59950583da013fc79e06db340"

Det commented on 2014-07-28 12:03

Removed already from -beta and -dev.

hadrons123 commented on 2014-07-28 05:16

libgcrypt dependency has been removed in chrome upstream. can you remove it from dependency list at the next release bump?

https://code.google.com/p/chromium/issues/detail?id=369973

Det commented on 2014-07-21 20:51

Fabulous, but next time preferably Pastebin.

weirddan455 commented on 2014-07-20 22:02

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 https://bugs.freedesktop.org/show_bug.cgi?id=38038
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.

CharlesAtum commented on 2014-07-18 17:46

This bug now also affects the Stable channel:
https://code.google.com/p/chromium/issues/detail?id=378090

Blake-R commented on 2014-07-18 04:53

After update to version (36.0.1985.125-1) there is a problem appears - all menus displays only on first screen (of two), even if chrome window - on second.

Det commented on 2014-07-10 19:04

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

...What don't you believe? Did setting XCURSOR_THEME not fix Chrome's cursor for you?

Det commented on 2014-07-10 18:54

I _don't_ believe you.

weirddan455 commented on 2014-07-10 16:01

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

Managed to find something in dmesg that seems correlated with the failures:
https://gist.github.com/catalin-hritcu/bd0a5b374a528b382cad
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

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

Done.

mityukov commented on 2014-06-23 19:11

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

Lol: "I am not planning on working on this soon. Putting back into the queue."

CharlesAtum commented on 2014-06-22 18:00

@catalin.hritcu
I reported the mouse cursor problem some time ago: https://code.google.com/p/chromium/issues/detail?id=356228
Looks like nobody is currently working to fix it.

catalin.hritcu commented on 2014-06-22 15:43

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?

hobarrera commented on 2014-06-18 06:46

@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

@Det: Thanks. Turned out I had a couple old desktop files in my home dir.

KevinCodux commented on 2014-06-10 19:13

New hashes:

i386: 2e51d1d06b6fdcb95b25042be0b44300
amd64: 0cf3cbd05f4a8c90e88b597b0c4028a0

KevinCodux commented on 2014-06-10 19:04

Current MD5/checksum is invalid

lockheed commented on 2014-06-10 15:05

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

└┌(%:~)┌- 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

matkam commented on 2014-06-10 07:01

The desktop shortcut that this creates does not work because it is pointing to the wrong binary. It points to "google-chrome" but the binary is "google-chrome-stable"

huorpalantir commented on 2014-06-04 07:44

Fonts and icons seem to be changed (smaller/larger/reverted to default) in Chrome 35 compared to Chrome 34. https://code.google.com/p/chromium/issues/detail?id=375824
Folks who value their font size or icons shouldn't update.

WoefulDerelict commented on 2014-06-04 02:16

Noted. Currently in the process of building a system off the standard ARCH image released this month to differentiate. I can find no major difference between Antergos and ARCH systems that would explain the error. The only change made to the systems between working install and failure was the update.

Det commented on 2014-06-04 02:04

Well, I did it for you: http://pastebin.com/JmnxEEGN

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.

WoefulDerelict commented on 2014-06-04 02:01

Pastebin would require me to be able to use a full featured brower on my system and on the afflicted systems said browser is unusable. The service is incompatible with Epiphany and resulted in errors when I attempted to load it; however, if more information is required the data can be moved to a system with working tools and uploaded to the transient pastembin and a link to the temporary content provided here instead of the actual data.

Det commented on 2014-06-04 01:49

Pastebin, please :D.

WoefulDerelict commented on 2014-06-04 01:48

In the event extra information can lead to solution of this issue the following is the log generated from a sync on a machine previously capable of running the chrome binary.[2014-06-03 15:01] [PACMAN] Running 'pacman -S --color never extra/ffmpeg-compat'
[2014-06-03 15:01] [PACMAN] installed ffmpeg-compat (1:0.10.12-2)
[2014-06-03 16:32] [PACMAN] Running 'pacman --color never -Sy'
[2014-06-03 16:32] [PACMAN] synchronizing package lists
[2014-06-03 16:32] [PACMAN] starting full system upgrade
[2014-06-03 16:33] [PACMAN] Running 'pacman --color never -S -u'
[2014-06-03 16:33] [PACMAN] starting full system upgrade
[2014-06-03 16:35] [PACMAN] upgraded clutter-gst (2.0.10-2 -> 2.0.12-1)
[2014-06-03 16:35] [PACMAN] upgraded e2fsprogs (1.42.9-2 -> 1.42.10-1)
[2014-06-03 16:35] [PACMAN] upgraded ffmpeg (1:2.2.2-3 -> 1:2.2.3-1)
[2014-06-03 16:35] [PACMAN] upgraded gdk-pixbuf2 (2.30.7-1 -> 2.30.8-1)
[2014-06-03 16:35] [PACMAN] upgraded gedit (3.12.1-1 -> 3.12.2-1)
[2014-06-03 16:35] [PACMAN] upgraded git (1.9.3-1 -> 2.0.0-1)
[2014-06-03 16:35] [PACMAN] upgraded gpgme (1.4.3-1 -> 1.5.0-1)
[2014-06-03 16:35] [PACMAN] upgraded gmime (2.6.20-1 -> 2.6.20-2)
[2014-06-03 16:35] [PACMAN] upgraded gnome-calculator (3.12.1-1 -> 3.12.2-1)
[2014-06-03 16:35] [PACMAN] upgraded gnome-icon-theme-symbolic (3.12.0-1 -> 3.12.0-2)
[2014-06-03 16:35] [PACMAN] upgraded gnome-icon-theme (3.12.0-1 -> 3.12.0-2)
[2014-06-03 16:35] [PACMAN] upgraded gnome-menus (3.10.1-1 -> 3.10.1-2)
[2014-06-03 16:36] [PACMAN] upgraded gnome-tetravex (3.12.2-1 -> 3.12.3-1)
[2014-06-03 16:36] [PACMAN] upgraded gupnp (0.20.11-1 -> 0.20.12-1)
[2014-06-03 16:36] [PACMAN] upgraded python-setuptools (3.6-1 -> 4.0.1-1)
[2014-06-03 16:36] [PACMAN] upgraded ipython (2.0.0-1 -> 2.1.0-1)
[2014-06-03 16:36] [PACMAN] upgraded json-glib (1.0.0-1 -> 1.0.2-1)
[2014-06-03 16:36] [PACMAN] upgraded libatasmart (0.19-2 -> 0.19-3)
[2014-06-03 16:36] [PACMAN] upgraded libgdata (0.15.0-1 -> 0.15.1-1)
[2014-06-03 16:36] [PACMAN] upgraded libinput (0.1.0-1 -> 0.2.0-1)
[2014-06-03 16:36] [PACMAN] upgraded libmms (0.6.2-1 -> 0.6.4-1)
[2014-06-03 16:36] [PACMAN] upgraded libxcursor (1.1.14-1 -> 1.1.14-2)
[2014-06-03 16:36] [PACMAN] upgraded lightdm (1:1.10.1-1 -> 1:1.10.1-2)
[2014-06-03 16:36] [ALPM-SCRIPTLET] >>> Updating module dependencies. Please wait ...
[2014-06-03 16:36] [ALPM-SCRIPTLET] >>> Generating initial ramdisk, using mkinitcpio. Please wait...
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'default'
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux.img
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> Starting build: 3.14.5-1-ARCH
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [base]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [udev]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [autodetect]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [modconf]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [block]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [keyboard]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [keymap]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [resume]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [filesystems]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [fsck]
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> ERROR: file not found: `fsck.btrfs'
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> WARNING: No fsck helpers found. fsck will not be run on boot.
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> Generating module dependencies
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> Creating gzip initcpio image: /boot/initramfs-linux.img
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> WARNING: errors were encountered during the build. The image may not be complete.
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'fallback'
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-fallback.img -S autodetect
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> Starting build: 3.14.5-1-ARCH
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [base]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [udev]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [modconf]
[2014-06-03 16:36] [ALPM-SCRIPTLET] -> Running build hook: [block]
[2014-06-03 16:36] [ALPM-SCRIPTLET] ==> WARNING: Possibly missing firmware for module: aic94xx
[2014-06-03 16:37] [ALPM-SCRIPTLET] ==> WARNING: Possibly missing firmware for module: smsmdtv
[2014-06-03 16:37] [ALPM-SCRIPTLET] -> Running build hook: [keyboard]
[2014-06-03 16:37] [ALPM-SCRIPTLET] -> Running build hook: [keymap]
[2014-06-03 16:37] [ALPM-SCRIPTLET] -> Running build hook: [resume]
[2014-06-03 16:37] [ALPM-SCRIPTLET] -> Running build hook: [filesystems]
[2014-06-03 16:37] [ALPM-SCRIPTLET] -> Running build hook: [fsck]
[2014-06-03 16:37] [ALPM-SCRIPTLET] ==> Generating module dependencies
[2014-06-03 16:37] [ALPM-SCRIPTLET] ==> Creating gzip initcpio image: /boot/initramfs-linux-fallback.img
[2014-06-03 16:37] [ALPM-SCRIPTLET] ==> Image generation successful
[2014-06-03 16:37] [PACMAN] upgraded linux (3.14.4-1 -> 3.14.5-1)
[2014-06-03 16:37] [PACMAN] upgraded linux-headers (3.14.4-1 -> 3.14.5-1)
[2014-06-03 16:37] [PACMAN] upgraded rygel (0.22.1-1 -> 0.22.2-1)
[2014-06-03 16:37] [PACMAN] upgraded usb_modeswitch (2.1.1-1 -> 2.2.0-1)
[2014-06-03 16:37] [PACMAN] upgraded xorg-xcursorgen (1.0.5-2 -> 1.0.6-1)

WoefulDerelict commented on 2014-06-04 01:42

I don't mind at all. Here is the output from one of the updated systems thanks to a little help from Epiphany so I could just paste it directly.

[derelict@Rattenkonig chrome]$ ls -l /opt/google/chrome/libudev.so.0 /usr/lib/libudev*
lrwxrwxrwx 1 root root 21 Jun 3 18:01 /opt/google/chrome/libudev.so.0 -> /usr/lib/libudev.so.1
lrwxrwxrwx 1 root root 16 Apr 13 00:11 /usr/lib/libudev.so -> libudev.so.1.4.0
lrwxrwxrwx 1 root root 16 Apr 13 00:11 /usr/lib/libudev.so.1 -> libudev.so.1.4.0
-rwxr-xr-x 1 root root 59960 Apr 13 00:11 /usr/lib/libudev.so.1.4.0

Det commented on 2014-06-04 01:35

I'm glad you're being helpful, but that's pretty obvious :).

CharlesAtum commented on 2014-06-04 01:34

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

Since this is obviously important to you, would you mind running:

$ ls -l /opt/google/chrome/libudev.so.0 /usr/lib/libudev*

WoefulDerelict commented on 2014-06-04 01:25

I would be much more willing to accept that as a valid response if the error wasn't so easily reproducible with any build of chrome or chromium not compiled against the latest kernel and libudev available in the core and community repositories. The updated packages are not unique to Antergos. The Antergos spin of ARCH utilises very few proprietary packages to generate the system, the majority of which are packages installing themes for GNOME and lightdm. No Antergos packages were updated during my recent sync; however, the kernel and usb_modeswitch packages were. Immediately after the update all builds of the chromium codebase compiled against older versions of the libudev library fail to launch of the system. All of the necessary files are in place and the symlinks are all sound; however, the chrome binary fails complaining it is unable to find libudev.so.0 where chromium runs into messier errors and terminates in a SEG_Fault

As both this and the beta package you maintain are debian packages and equally incompatible it is likely they were built against an older version of the kernel and library and are not compatible with the latest release pushed out to the ARCH community. I am unsure of what changes were made in the most recent updates but, I am able to create multiple ARCH spins based on the most current packages available and none will successfully launch a chrome or chromium binary.

Det commented on 2014-06-04 00:50

Then that's a problem with Antergos. Not with this.

WoefulDerelict commented on 2014-06-04 00:42

I am able to reproduce the error on multiple Antergos systems after applying the latest round of updates including usb_modeswitch2.2.0-1 and linux 3.14.5-1 Systems running with usb_modeswitch2.1.1-1 do not have the same libudev.so.0 error.

Det commented on 2014-06-04 00:30

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.

WoefulDerelict commented on 2014-06-03 23:59

Package is broken by updating to usb_modeswitch2.2.0-1. Chrome exits with an error stating it is unable to find libudev.so.O

Package flagged out of date.

WoefulDerelict commented on 2014-06-03 23:58

Package is broken by updating to usb_modeswitch2.2.0-1. Chrome exits with an error stating it is unable to find libudev.so.O

starnostar commented on 2014-06-03 00:20

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

@nicoazrak, done.

Det commented on 2014-05-31 00:54

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: http://www.nvidia.com/download/driverResults.aspx/76278/en-us

Det commented on 2014-05-30 20:21

At least on Gnome Shell the only thing I've noticed is the graphical corruption and sluggishness when resizing the window.

victorz commented on 2014-05-30 18:57

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! (crbug.com)

heaven commented on 2014-05-30 17:29

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

Update: v34 is a flag on top of the PKGBUILD for now.

krix commented on 2014-05-29 19:07

@Det Thank You very Much for 34 PKGBUILD! :) pipelight works great now.

Det commented on 2014-05-27 09:59

@max, http://blog.chromium.org/2013/09/saying-goodbye-to-our-old-friend-npapi.html

max commented on 2014-05-27 07:48

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

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: http://en.eduard-dopler.de/148/problems-with-chrome-icon-in-docky/

ddracopo commented on 2014-05-24 17:29

For me, v35 seems to have solved the previous problems of leaving still pages rendered on the screen when minimising the browser.

Det commented on 2014-05-23 19:48

And, as also mentioned, at least one possibility is your passwords.

catalin.hritcu commented on 2014-05-23 19:08

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

Bumping v34 for anyone else not reading the comments: http://pastebin.com/QEPxKwau

catalin.hritcu commented on 2014-05-23 18:55

I get strange graphics artifacts starting with Version 35.0.1916.114.

reed1 commented on 2014-05-22 01:50

thx Det! v35 was very awful, SmoothScroll also did not work.

Det commented on 2014-05-21 18:52

Those with LastPass don't have these kinds of issues :).

heaven commented on 2014-05-21 18:50

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

@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

What we've been talking about the whole time?

codenamestif commented on 2014-05-21 15:21

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

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

I just gave you the link for the PKGBUILD?

heaven commented on 2014-05-21 11:13

@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

The previous version wouldn't help you, because there's only one link for each channel, but: http://pastebin.com/QEPxKwau

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: http://pkgbuild.com/git/aur-mirror.git/tree/google-chrome, http://pkgbuild.com/git/aur-mirror.git/log/google-chrome

heaven commented on 2014-05-21 05:48

@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

@lucaskatayama, anyone? _Everyone_: http://blog.chromium.org/2013/09/saying-goodbye-to-our-old-friend-npapi.html

cellsheet commented on 2014-05-21 00:03

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

It's not finding java plugin. Anyone with the same problem?

Det commented on 2014-05-20 19:51

└┌(%:~)┌- google-chrome-stable
/usr/bin/google-chrome-stable: error while loading shared libraries: libgtk-x11-2.0.so.0: cannot open shared object file: No such file or directory

I disagree with you.

Det commented on 2014-05-20 19:45

What if I refuse?

skybon commented on 2014-05-20 19:43

Please remove gtk2 dependency: Chrome uses Aura by default now.

mssun commented on 2014-05-13 06:10

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

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?

veluria commented on 2014-05-11 14:22

@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

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

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

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

Det: we are all human!

Det commented on 2014-04-25 08:45

Broke it pretty quickly too.

mike.cloaked commented on 2014-04-25 08:44

Thank you for fixing the script so quickly.

Det commented on 2014-04-25 08:02

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

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

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

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

/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

@dmoebius, sounded like you just did.

dmoebius commented on 2014-04-21 19:01

Heh. Who cares about EULA anyways?!? Thx!

Det commented on 2014-04-21 17:11

Taken care of.

dmoebius commented on 2014-04-21 17:04

Yes, just now.

Det commented on 2014-04-21 16:18

Did you ever redownload it?

dmoebius commented on 2014-04-21 14:00

Since half a year or so, the md5sum for eula.txt is 6d57da7476a4b1b7a81821d9c036425c, not ba19ea498f294975d320ff0b26a6cd63. When will this be fixed?

dmoebius commented on 2014-04-21 13:58

Since half a year or so, the md5sum for eula.txt is 6d57da7476a4b1b7a81821d9c036425c, not 1f8d53940bf199d51c809f929e72543c. When will this be fixed?

Det commented on 2014-03-28 15:39

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')

..as 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

libcups should be a depends.

Det commented on 2014-03-04 13:51

Why? Are you from the future?

ushi commented on 2014-03-04 13:50

Integrity check failed for google-chrome-stable_33.0.1750.146_amd64.deb

Det commented on 2014-03-03 22:23

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

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

welp, sometime I see this ERROR:
ERROR:render_sandbox_host_linux.cc(643)] sendmsg: Broken pipe

heliumhgy commented on 2014-03-02 03:38

I have CPU usage issue. Chrome renders way slow and eats up too much CPU.
for test: http://www.douban.com/ (or even http://www.smzdm.com 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

Removed "PKGEXT='.pkg.tar'" due to a request of having your own way through makepkg.conf.

Det commented on 2014-02-23 17:42

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

Fix it then pastebin it, help the Maintainer. Or install it yourself.

Det commented on 2014-02-20 23:22

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

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

Only seconds after I looked it was already fixed... thank you Det (I presume).

mike.cloaked commented on 2014-02-20 21:53

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

I can't believe you guys raise such a fuzz about an outdated PKGBUILD.

rikless commented on 2014-02-20 19:47

concerns by FadeMind issue too.

FadeMind commented on 2014-02-20 19:10

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.

step-2 commented on 2014-02-20 18:50

the stable channel is at 33.0.1750.117 ??!

keepitsimpleengr commented on 2014-02-20 17:58

anybody know the md5sums for google-chrome 32bit ?

silent commented on 2014-02-20 17:29

md5sums for google-chrome amd64 Version 33.0.1750.117 should be 'a5c986f948ca72db015236fb9b6bb986'

heliumhgy commented on 2014-02-19 13:21

I get this error if i launch chrome in terminal:

ERROR:audio_manager_base.cc(422)] Not implemented reached in virtual std::string media::AudioManagerBase::GetDefaultOutputDeviceID()

and high CPU usage if I open multiple new pages at the same time and pages wont load. Those pages only load after the ERROR pop up in terminal, after that, CPU usage goes normal and everything becomes smooth.

Det commented on 2014-02-14 19:05

Removing 'libcanberra-pulse' was reported a long time ago to fix the issue: https://bbs.archlinux.org/viewtopic.php?pid=1228558

I don't know, if PulseAudio needs to be removed too.

reed1 commented on 2014-02-14 17:45

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

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

Anyone got random crash / segment fault? I got it far too often for a browser :( disabling extension / deleting profile folder won't help...

reed1 commented on 2014-02-08 16:06

Anyone got random crash / segment fault? I got it far too often for a browser :(

bartos commented on 2014-02-04 01:24

Working fine now 5 hours later. Sorry

Det commented on 2014-02-03 20:26

Works fine here.

bartos commented on 2014-02-03 20:01

curl: (22) The requested URL returned error: 404 Not Found
==> ERROR: Failure while downloading google-chrome-_32.0.1700.102_amd64.deb

Det commented on 2014-01-31 14:14

Remove /usr/lib/libgcrypt.so.11 (or /opt/google/chrome/libgcrypt.so.11) first.

Chrome doesn't use the 1.6 you get from the repos (libgcrypt.so.20).

leduck commented on 2014-01-31 08:44

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

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?

trippelganger commented on 2014-01-17 14:42

Is there a way to have both libgcrypt15 and 16 at the same time?

netsurfer commented on 2014-01-16 14:43

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

trippelganger commented on 2014-01-16 14:33

How do I use the patch? I tried replacing the PKGBUILD with it, but that just gave me a bunch of syntax errors!

ryuzy commented on 2014-01-16 10:45

With google chrome 32 I'm receiving Segmentation fault at the start.
Am I the only one?

$ google-chrome-stable
Gkr-Message: secret service operation failed: The name org.freedesktop.secrets was not provided by any .service files
Segmentation fault (core dumped)

Det commented on 2014-01-16 10:23

We're awesome, aren't we?

justinkb commented on 2014-01-16 10:01

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

I love you guys always handing out these patches to change a couple of lines :).

justinkb commented on 2014-01-15 23:36

http://bpaste.net/show/168513/

Simple patch to use the libgcrypt15 as suggested below

justinkb commented on 2014-01-15 23:35

http://bpaste.net/show/168513/

Simple patch to use the libgcrypt15 as suggested below

Will still need a lib32-libgcrypt15 package to support i386

r2b2 commented on 2014-01-15 20:16

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

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

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?

stmc commented on 2014-01-14 23:05

/usr/bin/google-chrome-stable: error while loading shared libraries: libgcrypt.so.11: cannot open shared object file: No such file or directory

lrm commented on 2014-01-14 20:51

The deb fails validation on 64-bit build. New checksum is: 5eda4b45f5482e526252a819ed5e78a3

lrm commented on 2014-01-14 20:47

The deb fails validation. New checksum is: 5eda4b45f5482e526252a819ed5e78a3

w1ntermute commented on 2014-01-14 20:13

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

And if you must know, pkgrel should have been incremented

nDray commented on 2014-01-14 20:07

New version is out, just put people out of their misery with the update.

Scimmia commented on 2014-01-14 17:44

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

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

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

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

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

you could increase the pkgrel number. this is exactly the use case.

Det commented on 2014-01-13 21:37

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

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

@jz_train just reinstall the package, also gets you libgcrypt.so.11. for a proper fix you have to wait for chromium/google to fix this

Det commented on 2014-01-13 20:16

And it's not out-of-date because of it..

Det commented on 2014-01-13 20:16

Or if you rebuild this thing.

jz_train commented on 2014-01-13 20:16

error while loading shared libraries: libgcrypt.so.11: 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

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

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

Even, if you redownload it? As in not using the one already in $startdir.

Det commented on 2014-01-01 00:33

Even, if you redownload it? As in not using the one already in $srcdir.

a1russell commented on 2013-12-30 07:44

eula_txt checksum is for me: 6d57da7476a4b1b7a81821d9c036425c

Det commented on 2013-12-26 13:56

Fixed. The package will now pull the previous libgcrypt.so(.11.8) 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

Oh.

hadrons123 commented on 2013-12-26 11:17

gentoo has some fix for this post libgcrypt 1.6 update.

Det commented on 2013-12-26 11:16

Then how did you get in here? Liar.

hadrons123 commented on 2013-12-26 11:16

Doesn't start after new libgcrypt 1.6 update.

JoveYu commented on 2013-12-15 04:08

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

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

I have tried, but nothing changed!

Det commented on 2013-12-13 08:56

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

why chrome render pages like this http://imgur.com/6WLs5R9

is it chrome's bug? or my computer's?

reed1 commented on 2013-12-11 04:41

@nhasian, not really

for those who faces strange icon grouping issue with kde task manager, go here: http://forums.opensuse.org/english/get-technical-help-here/applications/492582-google-chrome-kde-quick-launcher-issue.html

Det commented on 2013-12-07 22:31

You.

tioammar commented on 2013-12-06 15:58

I got this error message.. ./opt/google/chrome/PepperFlash/libpepflashplayer.so: Lzma library error: Corrupted input data. Is that the problem with the pkg or my system?

nhasian commented on 2013-12-06 04:53

@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

Here a PKGBUILD, works for me http://pastebin.com/hJm8zX0Q
Just change the source chrome link and md5sum

na12 commented on 2013-12-04 15:20

I get "[1204/160811:ERROR:nacl_helper_linux.cc(233)] NaCl helper process running without a sandbox!
Most likely you need to configure your SUID sandbox correctly" error and chrome then crash.

na12 commented on 2013-12-04 14:42

I get "Running without the SUID sandbox" error and then chrome crashed. What is the problem?

a1russell commented on 2013-12-04 01:36

eula_text checksum is failing.

reed1 commented on 2013-12-03 09:35

Anyone notice the shorcut isn't working anymore on KDE's task manager ?

http://i.imgur.com/sXd2Rac.png

magicalChicken commented on 2013-11-25 18:08

Could you please add libxtst to the dependencies, chrome crashes without it.

Det commented on 2013-11-25 08:44

Did you also notice the first comment from top?

mumbel commented on 2013-11-25 04:09

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

Just came here to say that now that the "Google Chrome Introduction Book" (http://www.google.com/googlebooks/chrome/) 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: http://blog.chromium.org/.

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

New version out:
http://googlechromereleases.blogspot.se/
31.0.1650.57

dmoebius commented on 2013-11-14 09:36

eula changed again. The new md5 sum is '6d57da7476a4b1b7a81821d9c036425c'.

ruario commented on 2013-10-23 08:19

@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

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

+1 yeah :D.

If nothing else, glad to see new users acquainting themselves with PKGBUILDs.

petko10 commented on 2013-10-17 21:33

+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

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

Weird, just updated google-chrome and eula's MD5 is 7ae235c61533ba1108b03d42c49e00bd /cc @Cameri

Cameri commented on 2013-10-17 13:54

The eula changed. New md5sum: a3a49767c80e92cca63b7a33aadfb2af

beststone commented on 2013-10-17 07:40

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

eula_text.html ... FAILED

joe1984 commented on 2013-10-16 22:00

I'm getting checksum a3a49767c80e92cca63b7a33aadfb2af for the file eula_text.html. Please fix this PKGBUILD.

MindfulMonk commented on 2013-10-03 11:34

Just change the PKGBUILD line

[ "$CARCH" = 'x86_64' ] && md5sums[0]='e4ab342703de87ec117cfea3951acce8'

to
[ "$CARCH" = 'x86_64' ] && md5sums[0]='7f055de3dc3a62cedb5156420c1bc99a'

thujone commented on 2013-10-02 09:33

On x86_64 I'm getting checksum: 7f055de3dc3a62cedb5156420c1bc99a

haaja commented on 2013-10-01 20:47

Just letting you know that there's an unnecessary .directory file in the tarball.

Det commented on 2013-09-20 18:26

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

@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 https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-amd64/Packages.gz | gzip -d | awk -F "[ -]" '/Package: google-chrome/ { print $4; getline ; print $2 }'

Det commented on 2013-08-09 16:26

Yeah, I've noticed that too back when I made that decision.

zatricky commented on 2013-08-06 15:02

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

I update it when I can.

step-2 commented on 2013-08-02 01:48

please update

lks commented on 2013-08-01 21:54

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

x64 new - md5sum 6d4367302896e62942184997770ddfa2

blackout24 commented on 2013-07-29 15:52

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

Chrome creates application shortcuts that look like this:

/opt/google/chrome/chrome --app=https://mail.google.com/mail/u/0

But they don't work:

$ /opt/google/chrome/chrome --app=https://mail.google.com/mail/u/0

/opt/google/chrome/chrome: error while loading shared libraries: libudev.so.0: cannot open shared object file: No such file or directory

The following does work:

$ /opt/google/chrome/google-chrome --app=https://mail.google.com/mail/u/0

Do I have to manually change my application shortcuts or can this be fixed in the PKGBUILD?

Anonymous comment on 2013-07-11 06:39

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

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: https://code.google.com/p/chromium/issues/detail?id=257998

TheWretched commented on 2013-07-08 14:48

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.

Anonymous comment on 2013-07-08 05:08

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

You can just flag it.

fiveofakind commented on 2013-07-03 17:45

v71 md5sums:

03de683b0e9b8bf77db17060c8113d4d google-chrome-stable_current_amd64.deb
19ec74343b5396760d5d6c88d7282cfb google-chrome-stable_current_i386.deb

fiveofakind commented on 2013-07-03 17:33

v70 x86_64 md5sum:
03de683b0e9b8bf77db17060c8113d4d google-chrome-stable_28.0.1500.70_amd64.deb

mike.cloaked commented on 2013-07-03 17:00

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

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

chrome always create "libpeerconnection.log" in my HOME

google say https://code.google.com/p/chromium/issues/detail?id=239048
maybe you can fix it for the moment
https://bbs.archlinux.org/viewtopic.php?id=163086

JoveYu commented on 2013-06-28 11:22

chrome always create "libpeerconnection.log" in may HOME

google say https://code.google.com/p/chromium/issues/detail?id=239048
maybe you can fix it for the moment
https://bbs.archlinux.org/viewtopic.php?id=163086

Det commented on 2013-06-26 20:11

Yeah, I really wish our libarchive was built against lrzip: https://bugs.archlinux.org/task/34655

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

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

Never ever heard of it (the software I mean).

I mentioned this to e36freak.

mariusn commented on 2013-06-26 09:37

I'm using cower and meat

Det commented on 2013-06-26 09:36

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?

Anonymous comment on 2013-06-26 08:24

@Det Seems to work like a charm - thanks ! :-)

Besides the zombie processes that is. Even after a full update including nvidia drivers. So stupid ...

Anonymous comment on 2013-06-26 08:13

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

==> 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

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.

Anonymous comment on 2013-06-22 13:03

The url format has also changed to:
"https://dl.google.com/linux/direct/google-chrome-${_channel}_current_${_arch}.rpm"

Anonymous comment on 2013-06-22 13:03

The url format has also changed to:
"https://dl.google.com/linux/direct/google-chrome-${_channel}_current_${_arch}}.rpm"

Anonymous comment on 2013-06-22 13:02

The url format has also changed to:
"https://dl.google.com/linux/direct/google-chrome-${_channel}_current}${_arch}}.rpm"

donniezazen commented on 2013-06-20 15:10

@Imaulle https://code.google.com/p/chromium/issues/detail?id=239048

Anonymous comment on 2013-06-20 13:37

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

Pacman has lots of hacks and it destroys binaries.

Anonymous comment on 2013-06-19 20:22

Why are there no zombie processes on other distro's ? Something in Arch has to interfere with Chrome...

Anonymous comment on 2013-06-19 17:06

still get zombie processes on 28.0.1500.45-205727 with nvidia driver. back to nouveau driver :(

Anonymous comment on 2013-06-19 14:27

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

@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 https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm | head -c96 | strings

colinkeenan commented on 2013-06-17 23:17

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.

Anonymous comment on 2013-06-16 19:03

A pity... because I need the nvidia driver for fast graphics. :-)

Anyway, killall /opt/google/chrome/chrome also works.

Anonymous comment on 2013-06-16 15:46

@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

Well I'm not the one that's gonna test it.

Anonymous comment on 2013-06-15 18:40

Replacing adblock by adblock+ didn't help ?

Anonymous comment on 2013-06-15 17:16

switching from nvidia driver to nouveau driver fixed the zombie processes for me

Det commented on 2013-06-12 08:20

&>/dev/null

adityarajbhatt commented on 2013-06-10 16:30

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 :

[741:766:0610/160010:ERROR:object_proxy.cc(624)] Failed to get name owner. Got org.freedesktop.DBus.Error.NameHasNoOwner: Could not get owner of name 'org.chromium.Mtpd': no such name
[741:766:0610/160010:ERROR:object_proxy.cc(624)] Failed to get name owner. Got org.freedesktop.DBus.Error.NameHasNoOwner: Could not get owner of name 'org.chromium.Mtpd': no such name
[741:741:0610/160010:ERROR:object_proxy.cc(529)] 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.

adityarajbhatt commented on 2013-06-10 10:33

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 :

[741:766:0610/160010:ERROR:object_proxy.cc(624)] Failed to get name owner. Got org.freedesktop.DBus.Error.NameHasNoOwner: Could not get owner of name 'org.chromium.Mtpd': no such name
[741:766:0610/160010:ERROR:object_proxy.cc(624)] Failed to get name owner. Got org.freedesktop.DBus.Error.NameHasNoOwner: Could not get owner of name 'org.chromium.Mtpd': no such name
[741:741:0610/160010:ERROR:object_proxy.cc(529)] 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

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

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

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: https://code.google.com/p/webrtc/issues/detail?id=1587

Anonymous comment on 2013-05-23 07:54

@t3ddy Impeccable as ever. Thanks ! :-)

mamamia88 commented on 2013-05-22 15:34

@mattijsf where did you find the verbld number?

netsurfer commented on 2013-05-22 13:40

@mattijsf i had to run "dos2unix PKGBUILD", then all ok! I dont now if its a my issue...probably not. thankyou

Anonymous comment on 2013-05-22 06:53

google-chrome 27.0.1453.93
http://pastebin.com/APvhhdEv

Updated PKGBUILD for the people who can't wait ;)

Det commented on 2013-05-22 05:26

I meant as in in here u cunt.

Anonymous comment on 2013-05-22 05:05

Only the 104th ? Damn... :-p

Det commented on 2013-05-22 04:05

Yes there is and you are like the 104th person who's noticed it.

cochise commented on 2013-05-22 03:52

There is a "always working" URL. I https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm.
What is need then is only change the md5 the new release.

mamamia88 commented on 2013-05-21 21:37

@nvm chrome shows correct version after rebooting the computer

mamamia88 commented on 2013-05-21 21:13

@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. https://docs.google.com/file/d/0B2DxWZFncgtPdXlxalozMHdkZWs/edit?usp=sharing

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

@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 http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog.html?url=/trunk/src&range=181864:190564&mode=html 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

mamamia88 commented on 2013-05-21 20:04

@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 http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog.html?url=/trunk/src&range=181864:190564&mode=html 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

Anonymous comment on 2013-05-21 18:54

aha ! chrome 27 has been released ?

weedfreak commented on 2013-05-19 09:35

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

I see you are also experiencing vision problems.

latelx commented on 2013-05-17 05:55

experice flash audio problem.

Det commented on 2013-05-14 13:15

No, you don't have to email him, because he's already getting notifications of our comments.

Anonymous comment on 2013-05-13 22:32

Thanks @carukia & @Det,

I've sent an email to the maintainer.

Det commented on 2013-05-13 17:58

@carukia, and I know it.

See: https://bugs.archlinux.org/task/27263, https://mailman.archlinux.org/pipermail/aur-general/2012-November/020859.html

Anonymous comment on 2013-05-13 17:26

@ryanpcmcquen I think only the maintainer can remove it.

Anonymous comment on 2013-05-13 16:51

Oops, accidentally flagged this out of date! Does anyone know how to remove the flag?

stuntgp2000 commented on 2013-05-10 13:16

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 https://code.google.com/p/chromium/issues/detail?id=224046 and it affects v26 and v27 beta

Anonymous comment on 2013-05-10 06:47

Processes are still remaining into memory. Shame...

Anonymous comment on 2013-05-08 19:47

New libpng is pushed, resulting in this:
$ google-chrome
/usr/bin/google-chrome: error while loading shared libraries: libpng15.so.15: cannot open shared object file: No such file or directory

michalmiddleton commented on 2013-04-26 19:12

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

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

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.

Anonymous comment on 2013-04-11 15:17

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

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

They like your memory. They don't wanna go.

You should be flattered.

Anonymous comment on 2013-04-11 07:46

Processes are stuck in memory again. Damn...

Det commented on 2013-04-09 18:40

I don't.

MrTux commented on 2013-04-09 18:10

source=("https://dl.google.com/linux/direct/google-chrome-${_channel}_current_${_arch}.rpm")

I think use this is better for update

Anonymous comment on 2013-04-08 21:47

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 (http://crbug.com/177218) 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

Let's hope 4.10.2 makes things work again.

Seems like the whole 4.10 series was a bit half-assed anyway: https://bugs.kde.org/show_bug.cgi?id=316086

Anonymous comment on 2013-04-04 06:48

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

Also have all those problems with sound and leftover processes.

gabrielrcp commented on 2013-04-02 04:09

ttf-font is provided by ttf-ms-fonts (amongst other true type fonts)
Relevant discussion:
https://mailman.archlinux.org/pipermail/arch-dev-public/2013-January/024252.html

Anonymous comment on 2013-04-02 02:38

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

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

Audio problem.
I solved starting with google-chrome %U --audio-buffer-size=2048

Anonymous comment on 2013-03-29 11:46

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

Just use either of these:

source=("https://dl.google.com/linux/direct/google-chrome-${_channel}_current_${_arch}.rpm")

or (debs: quicker)

elif [ "$CARCH" = "x86_64" ]; then
_arch='amd64'
[...]
source=("https://dl.google.com/linux/direct/google-chrome-${_channel}_current_${_arch}.deb")

..and build with "--skipinteg".

Anonymous comment on 2013-03-29 08:29

Newbie question: can the current make script be updated with the source & md5 indicated by k2s ?

Anonymous comment on 2013-03-28 08:53

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

current 64bit:
source=('https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm')
md5sums=('76524edb4ce9d3f946c050a48e831437')

Det commented on 2013-03-27 09:26

Yes: http://googlechromereleases.blogspot.fi/2013/03/stable-channel-update_26.html

Anonymous comment on 2013-03-27 09:25

Is it possible there was an update recently ?

pwalker commented on 2013-03-22 13:50

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 http://goo.gl/gbcAu for a comparison.

Det commented on 2013-03-14 12:35

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}_${_arch}.rpm"::"http(s)://dl.google.com/dl/linux/direct/google-chrome-${_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).

Anonymous comment on 2013-03-14 12:16

@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

@distronicon: http://googlechromereleases.blogspot.com/

Anonymous comment on 2013-03-13 18:56

@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

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

Why not using the link to the latest stable instead of building the url from some blog posting?

https://dl.google.com/linux/direct/google-chrome-stable_current_$arch.rpm

Where $arch is x86_64 or i386.

Anonymous comment on 2013-03-13 09:43

http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-stable-25.0.1364.160-186726.x86_64.rpm seems to be invalid, too :(

Det commented on 2013-03-11 17:58

I think it was due to Xorg 1.14.

TheWretched commented on 2013-03-11 17:53

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

The stable_current_x86_64 is already 25.0.1364.160.

Anonymous comment on 2013-03-08 03:18

The source link "http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-stable-25.0.1364.152-185281.x86_64.rpm" is invalid.
Use the following URL instead?
https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm

leduck commented on 2013-03-01 18:23

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

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

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

@ TheWretched

I agree this version seems to bee buggy under KDE 4.10.

TheWretched commented on 2013-02-28 20:09

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

You mean pacman -R libcanberra-pulse?

Anonymous comment on 2013-02-24 16:45

For those having issues with Chrome or Chromium freezing when something is typed in the address bar, try the fix outlined here: https://bbs.archlinux.org/viewtopic.php?pid=1228558

Det commented on 2013-02-18 10:25

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

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

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.

Anonymous comment on 2013-02-12 20:19

Please fix URL!

ckoch786 commented on 2013-02-12 16:28

Thank you for your contributions!

Anonymous comment on 2013-02-07 08:57

Works - thanks a lot for keeping it up to date ! :-)

t3ddy commented on 2013-02-06 21:14

When you get 404, just flag the package out-of-date and I'll update it ;)

Anonymous comment on 2013-02-06 20:17

It has been fixed, at least it doesn't throw a 404 anymore for me.

ingamedeo commented on 2013-02-06 19:50

Yes....404 error here! Please fix it!!

nickoe commented on 2013-02-06 19:43

Download link does currently throw a 404.

Det commented on 2013-02-06 18:14

No, it was kind of updated in between whenever it was.

mityukov commented on 2013-02-06 18:12

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

I changed the url to https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm to get it to work. The one provided in the comment on 01-31-2013 no longer is working (for me)

Anonymous comment on 2013-02-06 09:16

Worked like a charm. Thanks !

Anonymous comment on 2013-02-06 09:04

Worked like a charm. Thanks !

Anonymous comment on 2013-02-04 20:10

404 for me. bad link.

vendion commented on 2013-02-04 19:34

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

New URL for i386
http://dl.google.com/linux/chrome/rpm/stable/i386/google-chrome-stable-24.0.1312.57-178923.i386.rpm
MD5SUM = 2AE939E664AE538879F56D7C3C8C9970

New URL for x86_64
http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-stable-24.0.1312.57-178923.x86_64.rpm
MD5SUM = 6E374CBEB5314290E9F0C6586798B2E9

Anonymous comment on 2013-01-31 00:12

Had to modify the url in the PKGBUILD

Det commented on 2013-01-30 23:43

Wrong.

TheWretched commented on 2013-01-30 22:07

Package is broken, download source no longer exists with google.

jskier commented on 2013-01-28 20:54

I've been having issues with Chrome recently: https://bbs.archlinux.org/viewtopic.php?id=157048

Det commented on 2013-01-26 20:36

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

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

@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:

http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-stable-23.0.1271.97-171054.x86_64.rpm

The URL you pointing to works like a symlink to the actual build.

Anonymous comment on 2012-12-12 03:32

url scheme changed.. new url:

https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm

Anonymous comment on 2012-12-11 01:46

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/libpepflashplayer.so: truncated bzip2 input

dejavu commented on 2012-12-06 12:59

@Qoru: Thanks, enabling the switch "Override software rendering list" did the trick. Seems that my nvidia card was blacklisted recently.

Anonymous comment on 2012-12-03 14:17

Is it possible the older versions of chrome can not be downloaded from dl.google.com any more ?

mike.cloaked commented on 2012-11-30 09:52

It works nicely for me for version 23.0.1271.95-1

mike.cloaked commented on 2012-11-29 21:22

Another new version released today (29th November 2012)

Anonymous comment on 2012-11-27 19:11

Qoru's solution worked with Nvidia for me.

Det commented on 2012-11-27 18:41

Not with me it is.

Anonymous comment on 2012-11-27 18:30

Flash videos are playing but sound is distorted in this version.

Anonymous comment on 2012-11-23 19:58

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

madmack commented on 2012-11-20 01:40

Empty tab titles with the last update:

http://imageshack.us/a/img594/3171/emptytab.png

dejavu commented on 2012-11-19 13:21

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

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.

dejavu commented on 2012-11-19 00:32

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

Like the Intel/Catalyst drivers or some special kernel.

Willrandship commented on 2012-11-14 07:26

Well, I can confirm that fullscreen flash IS working...so it's not a general problem. It's something on your end.

rumpelsepp commented on 2012-11-12 17:46

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

Like f*ck it doesn't.

Anonymous comment on 2012-11-10 18:55

Confirmed, fullscreen flash video on youtube does not work with the latest update

stdigger commented on 2012-11-07 12:35

not work fullscreen flash video =(

stdigger commented on 2012-11-07 12:34

not work fullscreen flash video =(

Det commented on 2012-10-07 20:17

Yes :D. But it won't use it.

step-2 commented on 2012-10-07 20:16

can it be installed with gtk3 ?

msx commented on 2012-09-26 00:13

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!

msx commented on 2012-09-26 00:12

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 is very fun and hackish!

Anonymous comment on 2012-09-09 22:13

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

You have a worse judgment! >:O

Anonymous comment on 2012-09-09 22:07

Hello! Although a symlink was created for libudev.so.0 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

http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-stable-21.0.1180.79-151411.x86_64.rpm

Det commented on 2012-08-14 18:23

That's why I flagged it.

mike.cloaked commented on 2012-08-14 18:18

... yet another new version - .79 is out now.

ruario commented on 2012-08-14 10:44

@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

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

@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- https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm | bsdtar -xf- -C rpm
$ wget -qO- https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb | 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- https://dl.google.com/linux/direct/google-chrome-stable_current_i386.rpm | 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- https://dl.google.com/linux/chrome/rpm/stable/i386/repodata/other.xml.gz | 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- https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | 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:

https://dl.google.com/linux/chrome/rpm/stable/${_arch}/google-chrome-${_channel}-${_verbld}.${_arch}.rpm

or

https://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-${_channel}/google-chrome-${_channel}_${_verbld}_${_arch}.deb

ruario commented on 2012-08-14 07:37

Actually, on second thought you can make that APT query line simpler:

wget -qO- https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | gzip -d | awk '/Package: google-chrome-stable/ { getline ; print $2 }'

ruario commented on 2012-08-14 07:32

@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- https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm | bsdtar -xf- -C rpm
$ wget -qO- https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb | 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 icorey 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- https://dl.google.com/linux/direct/google-chrome-stable_current_i386.rpm | 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- https://dl.google.com/linux/chrome/rpm/stable/i386/repodata/other.xml.gz | 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- https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | gzip -d | awk -F "[ -]" '/Package: google-chrome-stable/ { getline ; print $2 "-" $3 }'
21.0.1180.77-r150576

The format of the version specific URLs is as follows:

https://dl.google.com/linux/chrome/rpm/stable/${_arch}/google-chrome-${_channel}-${_verbld}.${_arch}.rpm

or

https://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-${_channel}/google-chrome-${_channel}_${_verbld}_${_arch}.deb

mike.cloaked commented on 2012-08-13 21:15

Seems that version 21.0.1180.77 was pushed out rather soon after the previous version!

mike.cloaked commented on 2012-08-13 21:02

Seems that version 21.0.1180.77 was pushed out rather soon after the previous version!

Det commented on 2012-08-12 10:37

Naturally.

Deb: http://pastebin.com/cP2rZPwm

Rpm: http://pastebin.com/6329UEjW

antihero commented on 2012-08-12 09:57

Det do you have an up-to-date PKGBUILD I could use?

Det commented on 2012-08-12 09:55

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.

Det commented on 2012-08-12 09:54

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).

tancrackers commented on 2012-08-12 02:02

Why not use the .deb packages instead? Isn't chrome optimized mostly for Ubuntu anyways?

Det commented on 2012-08-11 16:17

Yeah, as already stated.

donniezazen commented on 2012-08-11 15:47

Did you version number change? I suggested as forbidden404 posted. My version number is still 21.0.1180.57.

crimsonknave commented on 2012-08-11 14:57

DL URL that I could find that worked was
https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm which worked with the checksum forbidden404 posted (for x86_64)

Det commented on 2012-08-09 16:36

It changes on every new build. That's just the general one to which I assume the latest one is linked to.

Anonymous comment on 2012-08-09 15:23

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=("http://dl.google.com/linux/direct/google-chrome-${_channel}_current_${_arch}.rpm")

Anonymous comment on 2012-08-01 01:38

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

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.

Det commented on 2012-07-22 11:42

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.

fladam commented on 2012-07-21 16:50

Because of Disk Quota, sometimes need chrome over 500 Mb

aliasbody commented on 2012-07-14 22:35

Just a silly question... Why isn't this package on the [extra]'s repositorie ?

t3ddy commented on 2012-07-10 14:36

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

He wants a disk cache of 10 bytes.

t3ddy commented on 2012-07-10 13:41

May I ask you why?

fladam commented on 2012-07-10 12:05

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

Could it link to /usr/lib/libudev.so.1 instead?
1.0.1 is installed by systemd-tools in [core], but the version in testing installs 1.0.2 instead.
libudev.so.1 (or even libudev.so) would work on both.

ataraxia commented on 2012-06-28 22:52

This doesn't need to depend on libpng12 any longer, it's been updated to use the current libpng (1.5).

jackgu1988 commented on 2012-06-28 18:58

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

Okay, the problem was in

Shockwave Flash 11.3 r31
11.3.31.109
/opt/google/chrome/PepperFlash/libpepflashplayer.so

after turning pepper flash off - everything back to normal

konart commented on 2012-06-28 17:17

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

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

Anonymous comment on 2012-06-28 13:05

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.

jackgu1988 commented on 2012-06-28 12:48

Anyone else having issues playing fullscreen youtube videos after the last update?

vorbote commented on 2012-06-27 11:02

Cannot upgrade automatically with packer, because although the PKGBUILD has been updated to 20.0.1132.43, the AUR version hasn't.

Fumon commented on 2012-06-02 23:25

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

thanks @eric chrome work flawlessly again

t3ddy commented on 2012-06-02 14:07

I've added the workaround to the pkgbuild. You can remove the simlink to keep the system cleaner.

Anonymous comment on 2012-06-02 13:37

Thanks ericcoleman

msx commented on 2012-06-02 08:06

@ericcoleman: +1

msx commented on 2012-06-02 07:58

+1

Anonymous comment on 2012-06-01 23:16

Temporary workaround for jstoik1 and vanni971's problem (as well as others I'm sure)

$ sudo ln -s /usr/lib/libudev.so.1.0.1 /usr/lib/libudev.so.0

Anonymous comment on 2012-06-01 21:54

Same problem of jstoik1

jstoik1 commented on 2012-06-01 18:36

Following the switch from udev to systemd-tools, google-chrome fails to run: error while loading shared libraries: libudev.so.0: cannot open shared object file: No such file or directory

donniezazen commented on 2012-05-31 21:47

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

More precisely 19.0.1084.46

Anonymous comment on 2012-05-16 12:21

Chrome 19 is out!

Anonymous comment on 2012-05-05 19:21

Well I am not 64 bit :)

Det commented on 2012-05-01 11:36

*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: https://projects.archlinux.org/svntogit/community.git/tree/trunk/PERMISSION?h=packages/skype

But even if they did move this some day, I'm not sure they'd keep both.

Det commented on 2012-05-01 11:25

Can't. For the same reason as Google Earth can't. They are binary packages.

Anonymous comment on 2012-05-01 06:40

I think google chrome should be in the extra repository. I don't like Chromium.

Det commented on 2012-04-29 11:19

Because you are so 64-bit.

Anonymous comment on 2012-04-29 08:50

Why I don't have flash inbuilt for google chrome ?

alexcortes commented on 2012-04-13 16:31

Not working here without libpng12.

poshus27 commented on 2012-04-07 08:46

Thanks :)

t3ddy commented on 2012-04-07 08:33

fixed :)

poshus27 commented on 2012-04-07 08:10

I removed libpng12 get error /usr/bin/google-chrome: error while loading shared libraries: libpng12.so.0: cannot open shared object file: No such file or directory.

Anonymous comment on 2012-04-07 01:23

I'm getting a 404 error.

mike.cloaked commented on 2012-04-06 16:19

Any chance this can be updated to version 18.0.1025.151? Thanks

nDray commented on 2012-03-29 12:58

It's a flash problem. See this, for example: http://r3dux.org/2011/12/how-to-partially-workaround-adobe-flash-plugin-issues-on-linux/

cowai commented on 2012-03-29 12:55

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

Yeah. That usually happens when they update their software and remove the old one.

Svenw commented on 2012-03-28 17:53

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

my mistake, again :)

wellington commented on 2012-03-22 14:58

Don't running without libpng12

Anonymous comment on 2012-03-22 02:06

17.0.963.83 is available now.

Det commented on 2012-03-12 12:36

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

@t3ddy
wait, your PKGBUILD still depends on openssl098, not openssl-compatibility.

t3ddy commented on 2012-03-11 12:54

Fixed, thanks.

xduugu commented on 2012-03-11 09:27

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

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

[ Temporary ] Working PKGBUILD:
http://pastebin.com/8puzXpTg

Det commented on 2012-03-10 19:43

T3ddy just figured that if the Dev Chrome was using it, so should this.

herve commented on 2012-03-10 17:56

Hi,

Was libpng12 removed as a dependency?

$ google-chrome
/usr/bin/google-chrome: error while loading shared libraries: libpng12.so.0: cannot open shared object file: No such file or directory

Thanks.

Anonymous comment on 2012-03-10 13:16

libpng12[0] is still a dependency, it seems to have gone missing from the PKGBUILD.

[0]: https://aur.archlinux.org/packages.php?ID=33795

Det commented on 2012-03-09 14:02

Because it's out of date.

crabman commented on 2012-03-09 13:55

building fails with:
curl: (22) The requested URL returned error: 404

Det commented on 2012-03-08 20:35

Yes. Lucid doesn't use it.

mike.cloaked commented on 2012-03-08 20:06

Is there any reason it can't depend instead on the openssl 1.0 from core?
http://www.archlinux.org/packages/core/i686/openssl/

xduugu commented on 2012-03-08 14:31

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] https://aur.archlinux.org/packages.php?ID=36308
[2] https://aur.archlinux.org/packages.php?ID=40574

Det commented on 2012-03-07 18:56

Uhhhhhhh... no?

mike.cloaked commented on 2012-03-07 18:49

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

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

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: "https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm"

Seems to build just fine after that.

BasioMeusPuga commented on 2012-03-07 08:40

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: "https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm"

Seems to build just fine after that.

Anonymous comment on 2012-03-07 07:19

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

Just a quick one http://pastebin.com/CmEWiR43

Anonymous comment on 2012-03-07 02:41

@jarav spotify installed those.

Hexcles commented on 2012-03-07 00:18

Package out of date. File not found(404) on dl.google.com.

Anonymous comment on 2012-03-06 18:57

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/libcrypto.so.0.9.8 exists in filesystem
openssl-compatibility: /usr/lib/libssl.so.0.9.8 exists in filesystem
Errors occurred, no packages were upgraded.

mike.cloaked commented on 2012-03-05 22:43

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

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

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.

Det commented on 2012-03-05 22:01

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?

mike.cloaked commented on 2012-03-05 21:43

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 21:37

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?

mike.cloaked commented on 2012-03-05 20:40

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

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

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 https://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/primary.xml.gz |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

Anonymous comment on 2012-03-05 14:20

fixed PKGBUILD (new version): http://pastebin.com/4jymgTbY

dejavu commented on 2012-02-24 19:27

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

Chrome's flash plugin is available only for 32 bit's package.
You've got to install flashplugin if you're on 64 bit.

Jristz commented on 2012-02-24 07:18

@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

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

Thanks for the package!
But I can't see the flashplugin in "about:chrome" included.

mike.cloaked commented on 2012-02-16 14:13

That was quick - and thanks for updating.

mike.cloaked commented on 2012-02-16 11:08

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

Of course it doesn't. That's repackaging not building.

Vrtak-CZ commented on 2012-02-05 22:29

/usr/bin/google-chrome: error while loading shared libraries: libpng14.so.14: cannot open shared object file: No such file or directory

after http://www.archlinux.org/news/libpnglibtiff-rebuilds-move-from-testing/

rebuild does not help :-(

poshus27 commented on 2012-01-24 08:50

Build error. http://pastebin.com/k5PwrkRd

poshus27 commented on 2012-01-24 08:50

Build error. http://pastebin.com/k5PwrkRd

t3ddy commented on 2012-01-14 09:29

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

Would you?

t3ddy commented on 2012-01-13 22:15

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

@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

@t3ddy: actually it seems makepkg now opens rpms (using bsdtar) for you! ;)

ruario commented on 2012-01-09 13:00

@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/google-chrome.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=$(wget -qO- http://dl.google.com/linux/chrome/rpm/stable/$(uname -m | sed 's/i.86/i386/')/repodata/other.xml.gz | zcat | awk -F\" '/-stable/ { print $10"-"$12 }')

Thus making the direct download urls:

http://dl.google.com/linux/chrome/rpm/stable/i386/google-chrome-stable-${_verbld}.i386.rpm
http://dl.google.com/linux/chrome/rpm/stable/i386/google-chrome-stable-${_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

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

> 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

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

@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

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

@t3ddy: I suspect this is what you want your meta script to do:

echo http://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-stable/google-chrome-stable_$(curl -s https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | 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.

ruario commented on 2012-01-09 07:56

If you only want the version number:

curl -s https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | zcat | awk -F "[ -]" '/Package: google-chrome-stable/ { getline ; print $2 }'

Or both the version and the build revision number:

curl -s https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | zcat | awk -F "[ -]" '/Package: google-chrome-stable/ { getline ; print $2,$3 }'

ruario commented on 2012-01-09 07:53

> 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.

curl -s https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | zcat | awk '/Package: google-chrome-stable/ { getline ; print }'

ruario commented on 2012-01-09 07:52

> 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.

curl -s https://dl.google.com/linux/chrome/deb/dists/stable/main/binary-i386/Packages.gz | gzip -d | awk -F "[ -]" '/Package: google-chrome-stable/ { getline ; print }'

ksmanis commented on 2012-01-07 01:11

PLEASE add kdebase-kdialog as an optdepend! Took me ages to figure out which package was needed.

Anonymous comment on 2012-01-03 16:06

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

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

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.

Det commented on 2011-11-16 18:55

T3ddy is already updating this thing every time someone flags it. He's just talking about not needing to change the build number every time.

Anonymous comment on 2011-11-16 18:11

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

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 :)

Anonymous comment on 2011-11-12 17:51

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. http://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-stable/google-chrome-stable_15.0.874.120-r108895_amd64.deb

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.

poshus27 commented on 2011-11-10 17:28

Add dependencies xdg-utils becouse not open downloads folder.

Det commented on 2011-10-27 10:41

Or you can just check the version number to see whether it matches with the latest one: http://googlechromereleases.blogspot.com/search/label/Stable%20updates

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.

Anonymous comment on 2011-10-27 09:24

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

@massol, that's not a fix - it does the same thing. Clear out your package cache first.

Anonymous comment on 2011-10-21 15:22

if md5sum failed, then try yaourt -Syu

unkier commented on 2011-10-07 04:09

yep, md5sum failed

thesame commented on 2011-09-20 19:24

md5sum failed

nDray commented on 2011-09-05 09:00

makepkg -g >> PKGBUILD every new release
http://googlechromereleases.blogspot.com/search/label/Stable%20updates

kardia commented on 2011-09-04 23:45

md5sum failed

Anonymous comment on 2011-08-26 02:36

md5sum failed

t3ddy commented on 2011-08-25 15:31

It still does: mv -v "$pkgdir/opt/google/chrome/google-chrome.desktop" "$pkgdir/usr/share/applications"

canuckkat commented on 2011-08-25 14:15

Is it just me or did this package used to make a menu item?

nDray commented on 2011-08-09 22:37

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

md5sum at least for x86_64 is wrong

Anonymous comment on 2011-08-02 16:57

The latest stable version is now 13.0.782.107 ;)

duca commented on 2011-06-15 21:44

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

New md5sum for 12.0.742.100
409c1d33d38e09bf295f8e57e18d210b google-chrome-stable_current_amd64.deb

Anonymous comment on 2011-05-04 12:09

As pciutils came to [core], now you can remove libpci dependency.

Det commented on 2011-04-07 13:33

@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.

Det commented on 2011-04-07 13:32

@unknown, still, libpci is yet only provided by the [testing] pciutils.

unknown commented on 2011-04-07 09:47

libxtst is missing from the dependencies.
libpci is already provided by pciutils.

unknown commented on 2011-04-07 09:44

libxtst is missing from the deps.

Det commented on 2011-04-06 08:28

@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?: https://bugs.archlinux.org/task/23437:

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).

Anonymous comment on 2011-04-06 03:02

Please, libpci should be dropped. Thanks!

Anonymous comment on 2011-03-26 01:16

When pciutils now in testing enters core the libpci dependency should be dropped: https://aur.archlinux.org/packages.php?ID=38263

poshus27 commented on 2011-03-08 17:10

The fast update, thanks.

Anonymous comment on 2011-02-03 19:10

New checksum for x86 is 41e6f6bb93d6bf98ef06941f0ee10194 and version is 9.0.597.84.

Det commented on 2010-12-17 13:33

Right.. didn't I just say so?

Anonymous comment on 2010-12-17 09:14

last version is 8.0.552.224

Det commented on 2010-12-15 09:16

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.

Anonymous comment on 2010-12-08 03:32

Hi. May you please update this PKGBUILD? Thanks in advance...

master commented on 2010-11-01 19:36

Package fixing Google Chrome's crash on opening jpgs: http://aur.archlinux.org/packages.php?ID=42507

Det commented on 2010-10-31 13:09

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.

Anonymous comment on 2010-10-31 12:44

it need libjpeg6,but by default,my arch install libjpeg8,what should I do?

Det commented on 2010-10-25 16:45

Yeah, I already notified x-demon about it. He accidentally checked in "http://googlechromereleases.blogspot.com/" 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.

Anonymous comment on 2010-10-25 12:09

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

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).

Det commented on 2010-10-21 17:02

A lot of crashes should've be fixed with 7.0.517.41 - did you update the package? You can do this by just redownloading it (the package is auto-updating).

Det commented on 2010-10-21 15:18

A lot of crashes should be fixed with 7.0.517.41 - did you update the package by redownloading it (auto-updating)?

blackdeagle commented on 2010-10-20 18:21

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

==> 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

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.

Det commented on 2010-09-22 14:53

Because this is built for the buggy, Ubuntu favured libjpeg6.

Det commented on 2010-09-22 14:52

Maybe, theoretically, just perhaps because of the fact that this is built for the buggy, Ubuntu favured libjpeg6 :)?

Det commented on 2010-09-22 14:51

Maybe, theoretically, just perhaps because of the fact that this is built for the buggy, ubuntu favoured libjpeg6 :)?

Det commented on 2010-09-22 14:50

The fact that this is built for libjpeg6, just perhaps :)?

Anonymous comment on 2010-09-19 19:05

Then what's the reason of crasing while attaching *.jpg files?

Det commented on 2010-09-19 17:26

@elfaimer, libjpeg6 _doesn't_ conflict with [extra]'s libjpeg.

Anonymous comment on 2010-09-19 00:51

Yes, I mean .jpg files. I don't want to compile anything else, I use Firefox for that job. It is just annoying.

Anonymous comment on 2010-09-18 22:23

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.

Anonymous comment on 2010-09-18 14:02

By the way, is there a solution for attached images crash?

Det commented on 2010-09-18 06:32

[extra]'s chromium also depends on bzip and libevent.

Anonymous comment on 2010-09-17 23:49

I installed xdg-utils package and now I can make Chrome default browser :) Maybe, you should add it as dependency.

Anonymous comment on 2010-09-17 17:52

I can't make Google Chrome my default browser. Any idea?

poshus27 commented on 2010-09-10 08:28

Please update browser thanks.

Fabioamd87 commented on 2010-09-03 10:02

google 6 stable: http://arstechnica.com/open-source/news/2010/09/chrome-reaches-second-birthday-version-6-goes-stable.ars?utm_source=rss&utm_medium=rss&utm_campaign=rss

trusktr commented on 2010-08-25 13:59

@WickedOne hmmm, that's strange, mine had everything!

trusktr commented on 2010-08-25 13:58

This package should be moved to the official repos to be along side firefox, etc!

Anonymous comment on 2010-06-15 02:54

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.

Anonymous comment on 2010-06-11 15:42

Is libjpeg6 and libpng12 still needed? Chromium (same version) compiles against the newer versions...

print commented on 2010-05-26 01:40

Here's the official announcement stating that chrome for linux is stable: http://blog.chromium.org/2010/05/google-chrome-for-linux-goes-stable.html

Anonymous comment on 2010-05-25 20:28

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.