Package Details: google-chrome-dev 53.0.2785.21-1

Git Clone URL: https://aur.archlinux.org/google-chrome-dev.git (read-only)
Package Base: google-chrome-dev
Description: An attempt at creating a safer, faster, and more stable browser (Dev Channel)
Upstream URL: https://www.google.com/chrome
Licenses: custom:chrome
Provides: google-chrome, pepper-flash
Submitter: None
Maintainer: Det
Last Packager: Det
Votes: 594
Popularity: 3.886713
First Submitted: 2009-06-05 21:02
Last Updated: 2016-07-20 14:19

Required by (22)

Sources (3)

Pinned Comments

Det commented on 2016-03-09 05:02

✔ 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/Dev%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 }'

Latest Comments

Det commented on 2016-06-30 14:37

Don't post a comment about it, flag the package.

KSarv commented on 2016-06-30 05:51

>> checksum

>What about it? Because it's totally up-to-date.

Can't build it too because of checksums

Det commented on 2016-03-25 21:49

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 Your package google-chrome-dev [1] has been flagged out-of-date by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

> checksum

What about it? Because it's totally up-to-date.

Det commented on 2016-03-09 05:02

✔ 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/Dev%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 2016-02-10 07:03

*OBS!* 32-bit support for Google Chrome Dev Channel has been dropped. 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

Det commented on 2016-02-04 09:35

I haven't had that in a long time, if you mean the one where I think all 4 components, Nvidia/Xorg/KDE/Chrome were involved.

E: Might have also been caused by no longer force-enabling GPU acceleration in chrome://flags/:

- https://bbs.archlinux.org/viewtopic.php?id=198817
- http://linux.slashdot.org/story/14/03/04/1926233/google-wont-enable-chrome-video-acceleration-because-of-linux-gpu-bugs
- https://bugs.launchpad.net/ubuntu/+source/chromium-browser/+bug/1466670

frebib commented on 2016-02-04 09:32

It seems not. In which case it's probably my config or something otherwise someone would have mentioned it by now.
After a couple hours, my PC grinds to a halt shortly before Chrome crashing and freeing up several gigs of RAM. I suspect a buggy extension

Det commented on 2016-01-30 12:54

Try and see. No one's mentioned that yet.

frebib commented on 2016-01-30 12:53

Is the horrendous memory leak fixed in this version?

Det commented on 2016-01-22 02:11

Let's try this one.

zeked commented on 2016-01-22 02:07

apparently the pcbeta mirror isn't archiving at least the unstable since 15 December 2015.

Det commented on 2015-12-17 15:36

@marcoms, read the description.

marcoms commented on 2015-12-17 15:34

==> Validating source files with md5sums...
google-chrome-unstable_49.0.2587.3_amd64.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!

Det commented on 2015-10-27 12:21

Yeah, I don't speak Russian, so... :)

But if it's not there, then it was either removed on purpose, or you should report (or look for) a bug.

Localizator commented on 2015-10-27 11:54

Det, not this settings on version 48.0.2541.0 http://imgur.com/VgrniyU

Det commented on 2015-10-26 13:09

Enable it first: chrome://flags/#enable-md-settings

Localizator commented on 2015-10-26 11:59

chrome://md-settings/ not worked on version 48.0.2541.0!

Det commented on 2015-10-16 07:34

Read the package description, please. :D

We got a long way till somebody didn't notice.

oboedad55 commented on 2015-10-15 23:45

Latest version;
Validating source files with md5sums...
google-chrome-unstable_47.0.2526.16_amd64.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!

Det commented on 2015-09-25 05:30

Thanks.

alumni commented on 2015-09-24 16:25

In the .desktop file the WM class is wrongly set. It is supposed to be:

StartupWMClass=google-chrome-unstable

Notice that the first letter is a lower case G instead of upper case.

Replacing all three occurrences will remove the duplicate chrome icon in Plank/Docky.

jvybihal commented on 2015-09-22 08:53

Is anybody else experiencing extremely long time to load? (at least one minute). I started having this issue yesterday, don't know what happend yet.


Edit: okay, nothing wrong with chrome here, it was a keyring that was causing the troubles.

jvybihal commented on 2015-09-22 06:10

Is anybody else experiencing extremely long time to load? (at least one minute). I started having this issue yesterday, don't know what happend yet.

Det commented on 2015-09-05 20:18

Pastebin.

3V0LU710N commented on 2015-09-02 01:59

It seems there are some problems with the md5sums:

PKGBUID:
md5sums=('00541a2eaedf40c2217f17f9345aaa5c'
'b7e752f549b215ac77f284b6486794b6')
md5sums_i686=('f732602aa7427da97ef1bf3ad4d673ed')
md5sums_x86_64=('2fc404c0fdd90c45acb5ed4d2cf38877')

Downloaded with "makepkg -g"
md5sums=('00541a2eaedf40c2217f17f9345aaa5c'
'b7e752f549b215ac77f284b6486794b6')
md5sums_i686=('f732602aa7427da97ef1bf3ad4d673ed')
md5sums_x86_64=('d1abecdb85ea5f37236684f75a51fd96')

Regards

3V0LU710N commented on 2015-09-02 01:58

It seems there are some problems with the md5sums...

Regards

Det commented on 2015-08-29 12:58

Flag. It. Don't. Post. The. Makepkg. Output.

EgidioCaprino commented on 2015-08-29 09:49

==> ERROR: One or more files did not pass the validity check!

Det commented on 2015-08-28 14:50

Read the top of the PKGBUILD.

GreenRaccoon23 commented on 2015-08-28 14:34

I think the md5sums for the 64 bit deb file should be this. I downloaded it twice and got the same result both times:
077c5822311c81f5404efa9c178e1c84

GreenRaccoon23 commented on 2015-08-28 14:06

Are the md5sums wrong or is it my network connection?

anish commented on 2015-08-27 18:10

Saweet ! Sorry for a delayed response, I had notifications turned off and did not realize

Det commented on 2015-08-26 17:48

Wouldn't have had time to do this tomorrow with our gig of 8AM - 9PM, but there's now support for flags read directly from ~/.config/chrome-dev-flags.conf. Didn't bump the pkgrel, but already seemed to work.

We apparently had manual support for $CHROMIUM_USER_FLAGS before through package().

Det commented on 2015-08-24 18:35

What I can do is add a ~/.config/chrome-flags.conf like Chromium has.

Might also do well to have a separate one for all channels.

Det commented on 2015-08-22 14:25

I don't fancy that, no.

anish commented on 2015-08-20 08:24

Would it be possible to add --force-device-scale-factor=1 to all the Exec lines in the .desktop file ? It fixes this issue https://code.google.com/p/chromium/issues/detail?id=490964 (though I'm not sure of the side effects of that on the ppl running hidpi setups)

Det commented on 2015-08-19 15:32

No, I'll keep it like this forever.

cooprocks123e commented on 2015-08-19 15:32

Wow, you updated it in the time it took me to register an account. Awesome. Sorry, can I unflag it or something?

Det commented on 2015-08-19 15:28

@cooprocks123e, why?

Det commented on 2015-08-15 11:18

Why ask me?

alumni commented on 2015-08-15 11:18

I can no longer pin Google Chrome to Plank/Docky. Is the StartupWMClass correctly set to "Google-chrome-unstable" in the .desktop file?

Det commented on 2015-07-16 21:55

Well, actually, you posted the "new md5" comment at 21:23 (UTC) and that one at 21:28.

I updated the package 21:02.

atriix commented on 2015-07-16 21:28

Sorry accidental flag, package updated literally in between the page refresh.

atriix commented on 2015-07-16 21:23

md5 for google-chrome-unstable_45.0.2454.6_amd64.deb seem to be
md5sums_i686=('ac79e91c215c2087caf49b59ec4a9ab9')
md5sums_x86_64=('7972e66df48b4208ace0f7885e4f6e6e')

Det commented on 2015-07-16 20:59

From the PKGBUILD:

# Check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Dev%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 }'

and then *flag*.

ycny commented on 2015-07-16 20:58

New hash:
md5sums_x86_64=('7972e66df48b4208ace0f7885e4f6e6e')

Det commented on 2015-06-10 12:10

You go, girl. :)

ruario commented on 2015-06-10 12:08

@Det:Yes, I am subscribed but don't read every comment right away. That is me.

I live in Norway but I'm actually British (my wife and daughter are Norwegian and I speak a little). I worked for Opera Software for a number of years (hence the interest in browsers). Shortly, I will be working for another browser company. ;)

Det commented on 2015-05-22 17:35

From the PKGBUILD:

# Check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Dev%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 }'

and then *flag*.

jevonearth commented on 2015-05-21 18:51

Appears that the checksums for google-chrome-unstable_44.0.2403.4_amd64.deb do not match the PKGBUILD.

324e766a06d6c2044822d8595eff1c1f google-chrome-unstable_44.0.2403.4_amd64.deb

Det commented on 2015-05-20 05:28

ruario, are you still subscribed to this comment section? You are from Norway, and have over 2000 posts in Linux Questions?: http://www.linuxquestions.org/questions/slackware-14/google-chrome-icon-4175492647/

Det commented on 2015-05-06 08:13

I don't agree with you.

mischka commented on 2015-04-22 07:52

md5sum fails for the deb on amd64

munyamunya commented on 2015-04-17 10:51

44.0.2369.0-1 still does not load tabs from previous session until selected, but atleast it shows their icons. Rejoice!

jugs commented on 2015-04-16 23:43

@GreenRacoon23, I still need to force the scaling factor to 1 otherwise it's comically large.

GreenRaccoon23 commented on 2015-04-14 22:47

@jugs It works fine for me, but I'm also running a moderately fast computer. The only bug I have is that the panel icon doesn't show up, even though the icons and .desktop files seem to be set up correctly in the file system.

jvybihal commented on 2015-04-13 10:15

Yes, with the modifications described below. But i personaly feel, that chrome is not performing as good as it was before.

jugs commented on 2015-04-13 10:11

Has anyone got this giant browser working properly with XFCE?

GreenRaccoon23 commented on 2015-04-11 22:50

(Thanks for that lifesaver hendry.)
The newest version (as of 2015-04-11) does not need the options that hendry posted.

hendry commented on 2015-04-08 07:16

export MESA_GL_VERSION_OVERRIDE=3.3
export MESA_GLSL_VERSION_OVERRIDE=330
google-chrome-unstable --force-device-scale-factor=1

Are the options I am using to get 43.0.2357.2-1 working sanely. Argh.

jugs commented on 2015-04-07 19:59

New shaders unsupported by mesa, new hidpi, how is this making it to the dev branch. Clearly no one is testing this... Here we downgrade again...

jvybihal commented on 2015-04-07 10:42

@jac3km4 thanks, that helped. But the browser is still freaking huge, as described by jugs. Screenshot: http://i.imgur.com/4wUPQ2x.png

jac3km4 commented on 2015-04-07 10:28

The new release uses GLSL 1.5 shaders, which are not supported by Mesa. You can enable the support as follows:

export MESA_GL_VERSION_OVERRIDE=3.3
export MESA_GLSL_VERSION_OVERRIDE=330
google-chrome-unstable

I suggest adding it to a startup script.

jvybihal commented on 2015-04-07 07:26

Latest build (43.0.2357.2-1) seems to be very broken. I am just receiving black window, as does other people:
https://code.google.com/p/chromium/issues/detail?id=474283
http://googlechromereleases.blogspot.cz/2015/04/dev-channel-update.html


If you use -dev as daily driver, I suggest you not to update until this is fixed.

jvybihal commented on 2015-04-02 06:34

I just downgraded, in PKGBUILD there is link to older version.

jugs commented on 2015-04-02 00:22

@jvybihal: How do we fix it?

jvybihal commented on 2015-04-01 15:18

Haha, I just installed it. The size of the browser is just massive :D

jugs commented on 2015-04-01 11:12

Wow... With 43.0.2351.3 my browser is absolutely massive. It looks like they "fixed"/added HiDPI support in linux in this build? How do I make it go back to regular size?

Det commented on 2015-03-31 23:22

Also, if you replace the md5sums with "md5sums_x86_64=('SKIP')", or build with "makepkg --skipinteg (-si)", you don't need to wait for me to update it.

Remember to bump the pkgver (from the PKGBUILD):

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: 41.0.2272.101
unstable: 43.0.2351.3   ⇦ this one
beta: 42.0.2311.60

jdarnold commented on 2015-03-31 20:54

Yes. If there is a new version (either you notice it beforehand or the checksum fails), just hit the out-of-date link.

Californian commented on 2015-03-31 20:48

There was a new version released today. Is the proper use of the "flag package out-of-date" button to use it now? I would like to upgrade but am still experiencing the validity check problem and hope that an update to the package can resolve that.

jvybihal commented on 2015-03-26 09:39

Guys, is anyone else experiencing this problem? http://i.imgur.com/wUnQbnk.png

Chrome and Chromium is not respecting "size" attribute in input element(s). When I via dev tools define "width: some_number px;" to that element, it spreads, but otherwise its shrinked. I see this on many pages.

Det commented on 2015-03-19 01:38

Downgraded for some reason:

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 }'
unstable: 43.0.2327.5
beta: 42.0.2311.50
stable: 41.0.2272.89

No mention in the Release Blog: http://googlechromereleases.blogspot.com/search/label/Dev%20updates

Det commented on 2015-03-18 02:23

Try with the latest Dev.

GreenRaccoon23 commented on 2015-03-17 22:37

Is the panel/window icon not showing up for anyone else with the latest build? I'm using xfce and have tested across multiple icon themes with the same problem. I tried adding "Icon=google-chrome-unstable" to "/usr/share/applications/Google Chrome (unstable).desktop" without success. (Also, thanks for maintaining this package, Det.)

Det commented on 2015-03-05 21:48

Updated.

dflt commented on 2015-03-05 21:45

==> Validating source files with md5sums...
eula_text.html ... Passed
==> Validating source files with md5sums...
google-chrome-unstable_42.0.2311.15_amd64.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!
==> ERROR: Makepkg was unable to build google-chrome-dev.

captainju commented on 2015-02-18 08:49

@Det you're right, I totally missed that mention, sorry :'(

Det commented on 2015-02-17 09:47

@captainju, it's for Windows and Mac: http://googlechromereleases.blogspot.com/2015/02/dev-channel-update_12.html

Det commented on 2015-02-17 09:47

@captainju, it's for Windows and Mac: http://googlechromereleases.blogspot.fi/2015/02/dev-channel-update_12.html

Det commented on 2015-02-12 14:41

Thanks, but that's pretty much all invalid.

tajidinabd commented on 2015-02-12 14:38

google-chrome-dev E: Missing custom license directory (usr/share/licenses/google-chrome-dev)
google-chrome-dev W: Referenced library 'libwidevinecdm.so' is an uninstalled dependency
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/libwidevinecdmadapter.so') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/libffmpegsumo.so') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/nacl_helper_bootstrap') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/libwidevinecdm.so') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/nacl_irt_x86_64.nexe') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/chrome') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/chrome-sandbox') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/nacl_helper') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/PepperFlash/libpepflashplayer.so') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome-unstable/lib/libpeerconnection.so') outside of a valid path.
google-chrome-dev W: File (opt/google/chrome-unstable/chrome-sandbox) is setuid or setgid.
google-chrome-dev W: Dependency alsa-lib included but already satisfied
google-chrome-dev W: Dependency included and not needed ('flac')
google-chrome-dev W: Dependency harfbuzz included but already satisfied
google-chrome-dev W: Dependency included and not needed ('harfbuzz-icu')
google-chrome-dev W: Dependency hicolor-icon-theme included but already satisfied
google-chrome-dev W: Dependency included and not needed ('icu')
google-chrome-dev W: Dependency libpng included but already satisfied
google-chrome-dev W: Dependency libxtst included but already satisfied
google-chrome-dev W: Dependency included and not needed ('opus')
google-chrome-dev W: Dependency included and not needed ('snappy')
google-chrome-dev W: Dependency included and not needed ('speech-dispatcher')
google-chrome-dev W: Dependency included and not needed ('ttf-font')
google-chrome-dev W: Dependency included and not needed ('xdg-utils')

hendry commented on 2015-02-11 09:23

No, I don't want to install an extension. I found the answer:
xdg-mime default google-chrome-unstable.desktop application/pdf

Det commented on 2015-02-11 02:10

This was pretty straightforward to Google up, but if the PDF viewer is enabled in chrome://plugins (default), you can use an extension: http://www.howtogeek.com/howto/18948/view-docs-and-pdfs-directly-in-google-chrome/

hendry commented on 2015-02-11 02:06

HI there, I noticed sometimes Chrome downloads PDF and when I double click the finished download it launches GIMP of all things. How do I make it so that Chrome opens it in the browser? Many thanks,

Det commented on 2015-02-10 12:53

/usr/share/applications/google-chrome-unstable.desktop

Look again.

scellow commented on 2015-02-10 12:49

It doesn't install the desktop entry ...

Det commented on 2015-02-03 22:12

Currently, nothing is failing validity check.

wicast commented on 2015-01-30 11:14

For md5sum of google-chrome-unstable_current_amd64.deb is now c7618adf2abe52acb55473eeccb4fa1b,please update the PKGBUILD.

Tharensis commented on 2015-01-30 02:08

amd64.deb is failing validity check.

wilberfan commented on 2015-01-30 01:59

The _i386.deb is failing the validity check for me.

Det commented on 2015-01-13 10:23

You are right, it's that way around.

hagabaka commented on 2015-01-13 09:13

@Det: when you quote "$VARIABLE" like the package currently does, multiple words in the variable will be treated as one single argument. When you don't quote it, the variable can be expanded to multiple arguments, split by space. Chromium implements CHROMIUM_FLAGS this way too, as you can see the last line of /usr/bin/chromium says

exec /usr/lib/chromium/chromium $CHROMIUM_FLAGS "$@"

Det commented on 2015-01-13 09:04

@ruario, I finally checked, and both the download and decompression were slower than with .debs.

@hagabaka, very few people use CHROME_USER_FLAGS, and many would be unaware why their flags work with Chromium but not Chrome:

- https://www.google.com/search?&q="CHROME_USER_FLAGS" (19 results)
- https://www.google.com/search?&q="CHROMIUM_USER_FLAGS" (1460 results)

Also _with_ quotes is how you get the multiple arguments.

hagabaka commented on 2015-01-11 21:39

Actually, it should probably be

sed -i 's/ "$@"/ $CHROMIUM_USER_FLAGS "$@"/' "$pkgdir"/opt/google/chrome-$_channel/google-chrome-$_channel

Without quotes around $CHROMIUM_USER_FLAGS. This way it's possible to set multiple flags separated by space.

hagabaka commented on 2015-01-11 21:08

The subsititution to provide support for CHROMIUM_USER_FLAGS isn't right. It currently results in

--user-data-dir="$CHROME_USER_DATA_DIR""$CHROMIUM_USER_FLAGS" "$@"

So the value of CHROMIUM_USER_FLAGS is appended to the user data dir making Chrome use a new profile. The sed line in PKGBUILD should be modified to have a space before "$CHROMIUM_USER_FLAGS":

sed -i 's/ "$@"/ "$CHROMIUM_USER_FLAGS" "$@"/' "$pkgdir"/opt/google/chrome-$_channel/google-chrome-$_channel

Also, what about using CHROME_USER_FLAGS instead? Some flags, like --ppapi-flash-path= and --ppapi-flash-version=, aren't necessary for Chrome.

Det commented on 2015-01-09 13:13

That was actually something only the RPMs were missing.

I'll think about it. Thanks.

ruario commented on 2015-01-09 13:12

The rpms also have the chrome-sandbox setuid already, which again allows for another potential simplification to the PKGBUILD.

ruario commented on 2015-01-09 13:09

The dependencies of Chrome unstable have recently changed.

$ wget -q https://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-unstable-41.0.2267.0-1.x86_64.rpm
$ mkdir google-chrome-unstable-41.0.2267.0-x86_64
$ tail -c+40093 google-chrome-unstable-41.0.2267.0-1.x86_64.rpm | xz -d | (cd google-chrome-unstable-41.0.2267.0-x86_64; cpio --quiet -imd)
$ find google-chrome-unstable-41.0.2267.0-x86_64 -type f -exec file {} \; | grep 'LSB \(shared object\|executable\)' | cut -d: -f1 | xargs -n1 objdump -p | grep NEEDED | sort -u
NEEDED ld-linux-x86-64.so.2
NEEDED libasound.so.2
NEEDED libcairo.so.2
NEEDED libc.so.6
NEEDED libcups.so.2
NEEDED libdbus-1.so.3
NEEDED libdl.so.2
NEEDED libexpat.so.1
NEEDED libfontconfig.so.1
NEEDED libfreetype.so.6
NEEDED libgcc_s.so.1
NEEDED libgconf-2.so.4
NEEDED libgdk_pixbuf-2.0.so.0
NEEDED libgdk-x11-2.0.so.0
NEEDED libgio-2.0.so.0
NEEDED libglib-2.0.so.0
NEEDED libgobject-2.0.so.0
NEEDED libgtk-x11-2.0.so.0
NEEDED libm.so.6
NEEDED libnspr4.so
NEEDED libnss3.so
NEEDED libnssutil3.so
NEEDED libpango-1.0.so.0
NEEDED libpangocairo-1.0.so.0
NEEDED libplc4.so
NEEDED libpthread.so.0
NEEDED librt.so.1
NEEDED libsmime3.so
NEEDED libstdc++.so.6
NEEDED libwidevinecdm.so
NEEDED libX11.so.6
NEEDED libXcomposite.so.1
NEEDED libXcursor.so.1
NEEDED libXdamage.so.1
NEEDED libXext.so.6
NEEDED libXfixes.so.3
NEEDED libXi.so.6
NEEDED libXrandr.so.2
NEEDED libXrender.so.1
NEEDED libXss.so.1
NEEDED libXtst.so.6

As you can see it no longer depends on libudev.so.0, so you can update the PKGBUILD to remove the "Symlinking missing Udev lib" section.

P.S. Just a small "suggestion" (feel free to ignore). You may also note that the rpms now use xz compression and hence are of comparable size to the .deb packages.

$ wget -q https://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-unstable/google-chrome-unstable_41.0.2267.0-1_amd64.deb
$ ls -l google-chrome-unstable-41.0.2267.0-1.x86_64.rpm google-chrome-unstable_41.0.2267.0-1_amd64.deb
-rw-rw-r-- 1 ruario ruario 46762094 jan. 6 20:54 google-chrome-unstable_41.0.2267.0-1_amd64.deb
-rw-rw-r-- 1 ruario ruario 47524020 jan. 6 20:54 google-chrome-unstable-41.0.2267.0-1.x86_64.rpm

I mention this in case you want to reconsider using an .rpm as a source, rather than the .deb. As you know .rpms are auto-extracted by makepkg, thus allowing for a minor simplification to PKGBUILD.

Det commented on 2015-01-08 13:29

Old builds: http://mirror.pcbeta.com/google/chrome/deb/pool/main/g/google-chrome-unstable/

Kwpolska commented on 2015-01-08 13:26

Newest version crashes on startup.

Det commented on 2014-12-10 12:00

It's been a little clunky since the update, but seemed to get fixed pretty fast.

spaceback commented on 2014-12-10 07:13

Hi, just for the record - amd64 .deb package is not available for download (404) as of now...

Det commented on 2014-12-09 22:36

Gomenasai. A fix has been committed in my official google-chrome-dev 41.0.2243.0-1 package.

dfryer1193 commented on 2014-12-09 22:33

I keep getting a failure in package() when trying to extract data.tar.lzma. Upon further investigation, it looks like data.tar.lzma in src/ has been renamed to data.tar.xz. Renaming data.tar.xz to data.tar.lzma, or changing data.tar.lzma to data.tar.xz in the PKGBUILD seems to fix this.

jwundrak commented on 2014-12-09 22:31

Get this error, when I try to update with yaourt (to Version 41.0.2243.0-1):
bsdtar: Error opening archive: Failed to open 'data.tar.lzma'

Det commented on 2014-12-03 18:23

This version, by merely purely coincidentally, is reported to seemingly have exactly what seems to be 666 votes.

CharlesAtum commented on 2014-10-01 18:11

Just wanted to point this new feature hidden behind a flag: http://crbug.com/360372 .
We can now use the tab audio indicator and the context menu to mute or unmute a specific tab.

Det commented on 2014-09-29 21:43

@rainbowdashc, why is it out-of-date?

CharlesAtum commented on 2014-09-28 11:52

http://crbug.com

hendry commented on 2014-09-28 04:27

What's the correct URL for reporting a bug on this package?

I noticed capturing microphone doesn't work http://s.natalian.org/2014-09-28/1411878266_1364x748.png

Thanks!

Det commented on 2014-09-16 21:46

Fixed.

blackhole commented on 2014-09-16 21:20

==> ERROR: One or more files did not pass the validity check!

blackhole commented on 2014-09-16 21:20

==> ERROR: One or more files did not pass the validity check!

weirddan455 commented on 2014-09-13 06:30

There's a file "google-chrome-unstable_39.0.2150.5_amd64.deb.part.aria2" in the tarball that doesn't seem to be doing anything. Did you package that by mistake?

ryanvade commented on 2014-08-29 03:35

==> Retrieving sources...
-> Found google-chrome-unstable_38.0.2125.24_amd64.deb
==> Validating source files with md5sums...
google-chrome-unstable_38.0.2125.24_amd64.deb ... FAILED

Det commented on 2014-08-16 17:54

That's the direct link. They both work.

neuralfrequency commented on 2014-08-16 10:27

Looks like the link to the .deb pkg is bad.

s/`source=("google-chrome-${_channel}_${pkgver}_$_arch.deb::https://dl.google.com/linux/direct/google-chrome-${_channel}_current_$_arch.deb")`/`source=("google-chrome-${_channel}_${pkgver}_$_arch.deb::http://dl.google.com/linux/deb/pool/main/g/google-chrome-unstable/google-chrome-unstable_38.0.2121.3-1_amd64.deb")`/ and build just fine though.

Det commented on 2014-08-01 20:34

In which case, again, you can flag it.

captainju commented on 2014-08-01 06:53

Back up
http://googlechromereleases.blogspot.fr/2014/07/dev-channel-update_29.html
38.0.2107.3

surlykke commented on 2014-07-31 13:05

Ok - thanks.

Det commented on 2014-07-31 10:53

Apparently they downgraded: http://googlechromereleases.blogspot.com/2014/07/dev-channel-update_29.html

Det commented on 2014-07-31 10:53

Apparently they downgraded: http://googlechromereleases.blogspot.fi/2014/07/dev-channel-update_29.html

surlykke commented on 2014-07-31 08:47

I'm getting md5sum mismatch when building.
Calculating md5sum on the downloaded debfile:

md5sum google-chrome-unstable_38.0.2107.2_amd64.deb

gives: 675170ac9d66c260176e6b33089e1769
where PKGBUILD expects: ebf6d432a44bbe3587b53879a7d99120

jvybihal commented on 2014-07-30 11:11

Damn, I ment to, I guess I didn't *oups*

Det commented on 2014-07-30 10:42

Flag it, please.

jvybihal commented on 2014-07-30 06:44

google-chrome-unstable_38.0.2107.2-1 :-)

gat commented on 2014-07-27 04:12

Thanks guys!
CharlesAtum: No, I couldn't reproduce this on my box with 38.0.2101.0.

CharlesAtum commented on 2014-07-26 01:27

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

Det commented on 2014-07-24 18:55

It sure is lovely to see you guys so persistently using the Dev Channel, though.

You are the root of our community!

CharlesAtum commented on 2014-07-24 18:22

The aliasing bug is already fixed in nightlys. Just wait until the next Dev.

jvybihal commented on 2014-07-24 15:32

Fonts are not antialiased and no hinting. I can confirm that. http://i.imgur.com/hvOwz6V.png

gat commented on 2014-07-24 15:29

Is it just me or are the jagged fonts back in 38.0.2101.0? :/

Det commented on 2014-07-23 21:28

You or your scripts didn't?

jayflo commented on 2014-07-23 15:37

This is not directly related to the package, but can anyone tell me what creates the ~/.config/google-chrome-unstable-backup directory and why it is so large?

Det commented on 2014-07-18 21:06

Also, previous builds available in: http://95.31.35.30/chrome/pool/main/g/google-chrome-unstable/

jakebailey commented on 2014-07-18 20:24

Here's an older bug report that's been tagged: https://code.google.com/p/chromium/issues/detail?id=395014

CharlesAtum commented on 2014-07-18 19:56

@stjhimy Found this bug: https://code.google.com/p/chromium/issues/detail?id=395090

stjhimy commented on 2014-07-18 19:41

yes @CharlesAtum, same here, pretty bad fonts. Any fix so far? Thanks.

CharlesAtum commented on 2014-07-18 18:38

Just me or anyone else is having a really bad font rendering on 38?

Det commented on 2014-07-18 18:11

Flag *out-of-date*, no?

niklas.heer commented on 2014-07-18 12:56

The md5sum is not correct! It should be 80c622448b4381648ad7d6493036e703. Thanks! :)

NettoHikari commented on 2014-06-26 01:12

Install fails on a fresh system at "Adding support for CHROMIUM_USER_FLAGS", because file "google-chrome" could not be found. Removing that line from PKGBUILD makes it work, though.

Det commented on 2014-06-18 03:39

Updated to .3.

phaseburn commented on 2014-06-17 18:41

37.0.2054.2 released :-) MD5 is 9fd8e4425aab37e6c29c4111951718fb

Cheers!

Det commented on 2014-06-05 03:32

Apologies. It's been a mistake from my update script.

ron.stewart commented on 2014-06-05 02:43

Failing to build due to mismatch on the md5sum; similar to previous version.

ernetas commented on 2014-06-04 16:06

How about we add amd64? Google says it's faster and more stable than 32bit version. Direct link to the appropriate deb file: https://dl.google.com/linux/direct/google-chrome-unstable_current_amd64.deb

Det commented on 2014-06-04 02:05

And uploaded.

hexchain commented on 2014-06-04 02:01

37.0.2024.2 released.

ron.stewart commented on 2014-05-31 02:25

@Det: thanks for the quick fix.

Det commented on 2014-05-30 20:27

Right. It's by the way not necessary to use the "./" when getting md5sums on the current folder.

ron.stewart commented on 2014-05-30 17:19

Build is currently failing for me:

makepkg -s
==> Making package: google-chrome-dev 37.0.2017.2-1 (Fri May 30 11:11:12 MDT 2014)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
-> Found google-chrome-unstable_37.0.2017.2_i386.deb
==> Validating source files with md5sums...
google-chrome-unstable_37.0.2017.2_i386.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!


May be a problem with the expected md5sum?

md5sum ./google-chrome-unstable_37.0.2017.2_i386.deb
a7499880b918cc2679780ad0e15b39cb ./google-chrome-unstable_37.0.2017.2_i386.deb

hexchain commented on 2014-05-29 11:36

37.0.2017.2 is out.

hexchain commented on 2014-05-14 05:23

36.0.1985.2 is out.

Raymondcal commented on 2014-05-12 20:31

Hi!
I’m using chrome-dev with Gnome 3, and it can’t remember that I clicked on the button to set it as the default browser.
Do you know how to fix it? Thanks!

Det commented on 2014-05-11 17:55

It's an NPAPI plugin. Aura Chrome for Linux v35+ only supports Pepper: http://blog.chromium.org/2013/09/saying-goodbye-to-our-old-friend-npapi.html

SZoPer commented on 2014-05-11 16:54

Anyone else having problem with Java Plugin? (java version "1.8.0_05")
I've already tried creating symlinks to /opt/java/jre/lib/amd64/libnpjp2.so in both /opt/google/chrome/plugins and /opt/google/chrome, but to no avail. I've got this issue for months.

augenleet commented on 2014-05-07 17:55

Alright, thanks. :)

Det commented on 2014-05-07 17:09

Means a new one up.

augenleet commented on 2014-05-07 16:59

Getting this error:

==> Validating source files with md5sums...
google-chrome-unstable_36.0.1964.2_amd64.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!
==> ERROR: Makepkg was unable to build google-chrome-dev.

colinkeenan commented on 2014-04-25 18:28

OK - Got google-talkplugin to work by re-installing it.

CharlesAtum commented on 2014-04-25 18:27

Well, I have these two files listed as Google Talk:
/opt/google/chrome-unstable/pepper/libppgoogletalk.so
/opt/google/chrome-unstable/pepper/libppo1d.so

They are just symlinks:
wxrwxrwx 1 root root 52 Mar 28 09:28 libppgoogletalk.so -> ../../../../opt/google/talkplugin/libppgoogletalk.so
lrwxrwxrwx 1 root root 45 Mar 28 09:28 libppo1d.so -> ../../../../opt/google/talkplugin/libppo1d.so

And both files are from the google-talkplugin package:
/opt/google/talkplugin/libppgoogletalk.so pertence a google-talkplugin 5.2.4.0-3
/opt/google/talkplugin/libppo1d.so pertence a google-talkplugin 5.2.4.0-3

Also, the pepper folder they are in seems to also be part of google-talkplugin:
/opt/google/chrome-unstable/pepper pertence a google-talkplugin 5.2.4.0-3

I'm also having a problem where Chrome doesn't seem to get the information that it is the default browser and clicking to change it doesn't work.

colinkeenan commented on 2014-04-25 18:04

Google Talk doesn't show up in chrome://plugins. I have the AUR package installed. CharlesAtum, do you have Google Talk installed and does it show up in chrome://plugins?

Also, the "unstable" branding is missing when setting google-chrome-unstable as the default web browser from the google chrome settings page. The previous version did that correctly, but for this one to work I had to cp google-chrome-unstable.desktop google-chrome.desktop in ~/.local/share/xfce4/helpers/.

Also, in case anyone using Xfce doesn't know, you can get Xfce Preferred Applications to recognize google chrome as default by editing the desktop file in ~/.local/share/xfce4/helpers/ and copying the stuff Xfce added to the end of the file, beginning with TargetEnvironment=Unity and put it at the end of the [Desktop Entry] section.

CharlesAtum commented on 2014-04-25 16:10

This one is still Aura, they just don't say it because there's no more need to differentiate anymore. Also, not sure why Google Talk doesn't work for you: AFAIK, the AUR package also installs PPAPI versions.

colinkeenan commented on 2014-04-25 15:42

Even though 36.0.1951.5 isn't aura, it still doesn't support google-talk-plugin or silverlight. I'm glad we can now have both dev and stable installed at the same time.

colinkeenan commented on 2014-04-25 15:37

I haven't had any problems with the dev aura builds lately and am happy with them. I also have intel graphics. By the way, this latest release, 36.0.1951.5, is just a plain dev (no aura) so it will probably work for you.

GeneArch commented on 2014-04-11 11:56

Is anyone else having problems with the screen. All versions for me since 35.0.1883.0 dev aur (including 1933) - when I start them up and it tries to restore the TABs - the pages are not filling and there are black squares and partial fills on the screen.

I have intel graphics.

1833 is working totally fine but all later versions are completely unusable for some reason.

Any suggestions?

thanks

gene

Det commented on 2014-04-04 23:49

Or just keep out of Dev Channel.

colinkeenan commented on 2014-04-04 23:47

Just installed *17 and it's working great. No more crashing on google-account login.

phaseburn commented on 2014-04-04 19:05

The crashing is a chrome issue, and has been acknowledged and fixed up stream as of this morning. See http://googlechromereleases.blogspot.com/2014/04/dev-channel-update_4.html for details...

Just need to wait for an AUR package to incorporate those fixes when Det gets around to it, and we should be good to go...

colinkeenan commented on 2014-04-04 06:11

Yes - the stable channel is working fine. I can sign into my google account without issue.

I may try once again with the unstable tomorrow now that everything is working and unstable was fully uninstalled. If it still doesn't work, must be a chrome issue. The only hope I have that it wasn't a chrome issue is that google-chrome-dev was upgraded and had trouble while my testing install became unstable and so packages weren't installing fully/properly. Now that everything's back to normal, if a re-install of unstable still crashes when I log into my google account, I have no doubt it's a chrome issue.

Det commented on 2014-04-04 05:28

Seems like something odd happened with the Pacman database, though, it would've been a lot easier to just 're-uninstall' it and then remove the settings from ~/.config and ~/.cache. Assuming the package isn't corrupt the installing part works with: # pacman -U --force /var/cache/pacman/pkg/google-chrome-dev-35.0.1916.14-1-*.pkg.tar.xz

Also, if Chrome is crashing when signing in to sync, then that definitely sounds like a Chrome problem. You never mentioned, if the Stable Channel (google-chrome) works fine?

colinkeenan commented on 2014-04-04 04:50

sudo find / -name *chrome-unstable* gave me 46 lines, but removing all the icons gives me the following 13 lines. I will rm all that and assume it's fully uninstalled.

/opt/google/chrome-unstable
/opt/google/chrome-unstable/google-chrome-unstable
/usr/bin/google-chrome-unstable
/usr/share/man/man1/google-chrome-unstable.1.gz
/usr/share/licenses/google-chrome-unstable
/usr/share/menu/google-chrome-unstable.menu
/usr/share/applications/google-chrome-unstable.desktop
/usr/share/gnome-control-center/default-apps/google-chrome-unstable.xml
/usr/share/doc/google-chrome-unstable
/home/colin/.config/google-chrome-unstable
/home/colin/.local/share/xfce4/helpers/google-chrome-unstable.desktop
/home/colin/.local/share/Trash/info/chrome-unstable.trashinfo
/home/colin/.cache/google-chrome-unstable

colinkeenan commented on 2014-04-04 04:33

In case I'm not being clear, when I talk about log in, I'm referring to logging into my google account. Google Chrome unstable is only crashing when I log into my google account - any one of them.

I would like to totally remove google-chrome-unstable though, so where can I see a file list for it so I can remove all the files? pacman already thinks it's uninstalled, but it's still there and runs other than the crashing issue.

colinkeenan commented on 2014-04-04 04:26

Yes Det, those seem to be the issues I was having but not anymore since I downgraded from testing.

I am able to use google-chrome-stable and login no problem. However, something strange has happened. Even though pacman thinks google-chrome-unstable is not installed, it nevertheless runs fine as long as I don't try to log in. Even the .desktop file still exists and works.

What are all the files I would have to manually remove in order to fully uninstall google-chrome-unstable.

Det commented on 2014-04-04 04:25

So, for reference:
- The fixing commit ([testing]/linux 3.14-3): https://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/linux&id=2c69d1905dc88e27181b36d523c9fa17a0d564c4
- LKML thread: http://www.gossamer-threads.com/lists/linux/kernel/1887224
- Another one: http://www.mail-archive.com/linux-kernel@vger.kernel.org/msg617266.html

Det commented on 2014-04-04 04:10

Reference:
- [testing]/linux 3.14-3 commit: https://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/linux&id=2c69d1905dc88e27181b36d523c9fa17a0d564c4
- LKML thread: http://thread.gmane.org/gmane.linux.kernel/1666905
- Another one: http://www.spinics.net/lists/kernel/msg1716924.html

Det commented on 2014-04-04 03:57

Reference:
- testing/linux 3.14-3 commit: https://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/linux&id=2c69d1905dc88e27181b36d523c9fa17a0d564c4
- LKML thread: http://thread.gmane.org/gmane.linux.kernel/1666905

Det commented on 2014-04-04 03:55

Reference: https://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/linux&id=2c69d1905dc88e27181b36d523c9fa17a0d564c4, http://thread.gmane.org/gmane.linux.kernel/1666905

colinkeenan commented on 2014-04-04 03:42

I use Xfce and don't use gnome keyring, so no seahorse. However, basically everything I tried to re-install was having issues so with a lot of trouble (at one point boot was failing) I finally downgraded from testing.

I downgraded the kernel and everything else down so that nothing was from testing. I re-installed google-chrome-unstable and (unlike before) didn't get any errors during the install.

After all that though, google-chrome-unstable was still crashing with segfault. So, I did:
rm -r .config/google-chrome-unstable
rm -r .cache/google-chrome-unstable

It didn't crash until about half a minute after I logged in. So, I removed those folders again and tried logging in under a different account. Still crashed half a minute later. Now I'm using google-chrome-unstable no problem as long as I don't log in.

I will try installing google-chrome-stable and see if I can log in.

CharlesAtum commented on 2014-04-04 02:05

@colinkeenan I'm not sure about the segfault but that Gtk-message could be because of seahorse (if you are running gnome). Could you check if it's working or run chrome with "--password-store=basic"? Ignore this if you don't care.

About the kernel: Reinstall the old kernel, test Chrome and update again. If Chrome fails after that then maybe it is really in the kernel (The devs are trying to trim it down, they may have disabled something chrome needed)

colinkeenan commented on 2014-04-04 01:17

I downgraded to *6 from April 1, and am having the same issue.

It started after I tried a system upgrade that tried to upgrade testing/linux 3.14-2 2 -> 3
but failed due to the file not having the required signature.

Seems like my issue is not being caused by *14 though. I'll work on it and probably end up posting for help in the forums.

colinkeenan commented on 2014-04-04 00:57

Having trouble with *.14

colin ~$ google-chrome-unstable
[3229:3229:0403/193416:ERROR:component_loader.cc(138)] Failed to parse extension manifest.
ATTENTION: default value of option force_s3tc_enable overridden by environment.
[3229:3229:0403/193416:ERROR:desktop_window_tree_host_x11.cc(1266)] Not implemented reached in void views::DesktopWindowTreeHostX11::MapWindow(ui::WindowShowState)
Gkr-Message: secret service operation failed: The name org.freedesktop.secrets was not provided by any .service files
[72,2513151744:00:34:16.968253] Native Client module will be loaded at base address 0x0000367700000000
[3229:3345:0403/193416:ERROR:get_updates_processor.cc(214)] PostClientToServerMessage() failed during GetUpdates
Segmentation fault (core dumped)

rainbowdashc commented on 2014-04-03 17:35

Hmm, just tried to upgrade for x86_64 and failed to verify the md5sum (the file I received had a sum of 88cc0fa5a83acaef7f66a44bf6ae3b1d)

Update: Version bumped to *.14

rainbowdashc commented on 2014-04-03 17:03

Hmm, just tried to upgrade for x86_64 and failed to verify the md5sum (the file I received had a sum of 88cc0fa5a83acaef7f66a44bf6ae3b1d)

tante commented on 2014-03-28 09:24

Doesn't show the tray icon for me (Gnome 3)

oconnors11 commented on 2014-03-25 16:33

@CharlesAtum I am experiencing the same issue with the icon. Also, am I missing something for Google Now? I have it enabled in chrome://flags

Det commented on 2014-03-25 08:08

@danyf90, my computer crashed when I was trying to upload the package. Then, when I finally got it up and running the server rejected it, complaining something about it being too good.

danyf90 commented on 2014-03-24 16:59

Hi, all the files that this package installs have a suffix -unstable (either in folder name or in file name), so why is this in conflict with google-chrome?

Regards,
Daniele

colinkeenan commented on 2014-03-23 23:40

This version (35.0.1897.2) is "35.0.1897.2 dev aura", but even though an "aura" build, it's working well for me. It even recognizes middle-click paste which other "aura" versions didn't that I tried. It has google-talk-plugin, but no silverlight. I will just use firefox for silverlight from now on.

crossRT commented on 2014-03-19 06:23

main .deb file did not pass the validity check. Is it because of out of date?

CharlesAtum commented on 2014-03-12 13:04

Hmm, looks like the icons should also be called google-chrome-unstable.

CharlesAtum commented on 2014-03-12 03:16

The Chrome icon is now using the generic binary icon after last update, at least for me. Is it also happening to somebody else?

Det commented on 2014-03-11 20:22

They did, huh.

CharlesAtum commented on 2014-03-11 20:20

They renamed usr/share/man/man1/google-chrome.1 to google-chrome-unstable.1

jnbek commented on 2014-03-04 03:44

having problems with Copy/Pasting. If I highlight text, or try to paste into a form/urlbar etc. I freezes the current tab and renders it unusable.

CharlesAtum commented on 2014-03-02 15:05

@colinkeenan Actually they ship two PPAPI libs realated to Google Talk. According to chrome://plugins/ they are:
Google Talk Plugin (/opt/google/chrome-unstable/pepper/libppgoogletalk.so)
Google Talk Plugin Video Renderer (/opt/google/chrome-unstable/pepper/libppo1d.so)
So it should be working already.

Det commented on 2014-03-01 18:54

1) Not likely, since the library is included in the main binary itself.

Why would you need to know that beforehand? If you don't want to install it, you don't need to pass '-i' to makepkg. To install the previous version you can either simply pull it from /var/cache/pacman/pkg/ or have a look in http://mirror.pcbeta.com/google/chrome/deb/pool/main/g/ or http://95.31.35.30/chrome/pool/main/g/.

2) Pepper is Google's new plugin API (the only thing currently using it is their Flash). It's likely not going anywhere anytime soon. NPAPI is the old platform, currently in use by everything else and is unsupported in Aura. It's not a plugin-based support, it's an API-based one: https://groups.google.com/a/chromium.org/forum/#!msg/chromium-dev/xEbgvWE7wMk/D_07G2lftacJ, http://blog.chromium.org/2013/09/saying-goodbye-to-our-old-friend-npapi.html

3) It was a joke to nibble the peculiar statement of Google not paying attention to Linux anymore.

colinkeenan commented on 2014-03-01 16:36

Is there a way too see by examining the tarball or .deb file whether or not a particular release is aura before actually installing it?

Also, I made a mistake typing "...no pepperflash" because that's the one they will still support in aura. But, definitely no silverlight and I think google-talk-plugin won't be supported either.

Android?

Det commented on 2014-02-28 17:54

Don't say that. Those f**king guys won't take Android away from me.

colinkeenan commented on 2014-02-28 16:22

When that broken version hit the dev channel, I switched to beta. Now that version hit beta, I tried to switch back to dev, but it's an aura release meaning no silverlight or pepperflash. So, now I'm back to stable. I hope these releases never make it to stable, but it doesn't seem like Google is actually paying attention to Linux anymore.

Det commented on 2014-02-27 05:50

Are you perhaps trying to be funny?

As a non-Google employee and a non-Chrome Dev Channel Release Manager I have very _little_ power over them and their choices to publish broken builds.

Their servers only hold the latest version of each Channel so if it's broken, then that's what it bloody hell is.

If you're looking for a stable browser, then you probably want to stay the hell away from the Dev Channel anyway.

Then if you _insist_ and don't have those older builds in '/var/cache/pacman/pkg/' you may go look in places like: http://mirror.pcbeta.com/google/chrome/deb/pool/main/g/google-chrome-unstable/, http://95.31.35.30/chrome/pool/main/g/google-chrome-unstable/

hendry commented on 2014-02-27 03:07

Damn, 34.0.1847.14 still completely unstable :( https://twitter.com/kaihendry/status/438871189797076992

Can you please upload packages that kinda work?

Det commented on 2014-02-26 17:07

$ sudo sed -i "s/PKGEXT=.*/PKGEXT='.pkg.tar'/" /etc/makepkg.conf

And enjoy the whole AUR.

xvoprsalek commented on 2014-02-26 14:16

Now the ' -> Compressing package...' takes several times longer.

Det commented on 2014-02-24 05:16

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

phaseburn commented on 2014-02-23 23:24

I had the same issue as @wilberfan, with the renderer locking up constantly.

I nuked my whole Chrome profile, didn't sign in upon opening Chrome, went to gmail.com and signed in there (so it wouldn't sync my profile back), and it locked up with just that tab open.

Downgrading to 1838 resolved all issues of the renderer process crashing on gmail and other sites.

wilberfan commented on 2014-02-23 06:23

This build doesn't work well for me at all: repeated "Page(s) Unresponsive" errors. (Which is probably why I can't scroll any pages, either...)

Det commented on 2014-02-20 12:38

Those busy buggers.

CharlesAtum commented on 2014-02-20 12:20

According to http://googlechromereleases.blogspot.com.br/ there's a new Dev version available.

colinkeenan commented on 2014-02-19 17:42

Thanks for all your help, Det. I'm glad I asked. I followed your advice that you posted on my package's AUR page.

Det commented on 2014-02-19 14:21

Mm. The .deb you can extract with bsdtar (which is done automatically) and contains the following files (the data.tar.lzma is the only thing we care for):

control.tar.gz (6,5 kB)
data.tar.lzma (45,9 MB)
debian-binary (4 B)

It may also be something like .tar.gz or .tar.xz, but in Chrome's case it's a .lzma. Either way, this one you have to extract yourself (from the PKGBUILD):

msg2 "Extracting the data.tar.lzma"
bsdtar -xf data.tar.lzma -C "$pkgdir/"

So it's not like makepkg only loves me, rather, it just does what I'm telling it to do.

colinkeenan commented on 2014-02-19 14:00

I have a question about the PKGBUILD because I've uploaded a couple of packages (nautilus-data and gnome-search-tool-no-nautilus) to AUR for the first time and am mystified by how you do "Extracting the data.tar.lzma" even though you downloaded a .deb file and never got data.tar.lzma out of it. My package nautilus-data also downloads a .deb file, but I had to use deb2targz available in the AUR in order to use it.

rainbowdashc commented on 2014-02-14 02:15

34.0.1838.2 released.

hazey commented on 2014-02-06 00:38

Just making sure you were aware lol - and ignore comment on link not working direct I wasn't paying attention that it was renamed in PKGBUILD, thanks for maintaining.

hazey commented on 2014-02-06 00:37

-and ignore last comment on the can't download from link in PKGBUILD, wasn't paying attention that it's renamed in there. MD5 issue remains, checked from the direct link and it did change to that MD5 mentioned prior comment - must have rebuilt or the PKGBUILD md5 wasn't updated

Det commented on 2014-02-06 00:33

Calm down.

hazey commented on 2014-02-06 00:32

Build failing on md5sum check for the main deb file on x64??? MD5 seems to be 560687ebc0dedda214188916a629ad16 which is differ from the PKGBUILD, also weird I cannot seem to download the file direct using the src link in the PKGBUILD but not so big a deal, MD5SUM is what is important/odd. Any ideas? Thanks!

hazey commented on 2014-02-06 00:25

Build seems to be failing on x64 from invalid checksum?

==> Retrieving sources...
-> Downloading google-chrome-unstable_34.0.1809.0_amd64.deb...
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 46.5M 100 46.5M 0 0 9569k 0 0:00:04 0:00:04 --:--:-- 9.8M
-> Downloading eula_text.html...
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 52500 0 52500 0 0 245k 0 --:--:-- --:--:-- --:--:-- 246k
==> Validating source files with md5sums...
google-chrome-unstable_34.0.1809.0_amd64.deb ... FAILED
eula_text.html ... Passed
==> ERROR: One or more files did not pass the validity check!
==> ERROR: Makepkg was unable to build google-chrome-dev.

Det commented on 2014-01-31 19:14

You changed your mind?

rainbowdashc commented on 2014-01-31 18:52

There appears to be no issues when running the dev channel with libgcrypt from [core].

sudsbud commented on 2014-01-30 05:31

The aura version of chrome does not support NPAPI plugins.
https://groups.google.com/a/chromium.org/forum/#!msg/chromium-dev/xEbgvWE7wMk/D_07G2lftacJ
I'm keeping a version of chromium on my ignore upgrade list.

colinkeenan commented on 2014-01-30 02:42

Ever since upgrading to Version 34.0.1809.0 dev aura, my plugins from AUR aren't installed in Chrome anymore. When I look at chrome://plugins/, neither google-talkplugin nor silverlight are there. I tried forcing silverlight to re-install by deleting ~/.wine-pipelight, but that didn't work. It did force a re-install on Firefox when I ran it though.

What's going on? Anyone else loose their AUR installed Chrome plugins on this version of Chrome?

joren commented on 2014-01-21 23:04

new checksum:
555e5c2af6bae57d7e02a11eb0435d76 google-chrome-unstable_current_amd64.deb

Det commented on 2014-01-21 06:11

Because there _isn't_ one: https://www.archlinux.org/packages/core/x86_64/libgcrypt/

You mirror is either a bit out-of-date or you haven't tried to actually update your system (which seems a bit odd taken how you're using a Dev Channel Chrome anyway).

PocketSam commented on 2014-01-21 05:07

Det, there is
core/libgcrypt 1.5.3-1

why not to use it instead?

Det commented on 2014-01-18 04:43

Well, it isn't :D.

Removing it from the dependencies doens't mean it doesn't use it.

Try "# pacman -R libgcrypt15" and then running Chrome again. It won't.

RichardH commented on 2014-01-17 22:21

I removed libgcrypt15 from the depends, and
chrome34 is running just fine with libgcrypt 1.6.0-1 from core.

sasy360 commented on 2014-01-17 04:20

With 34 out, Shouldn't it be on Aura by now?

DiFuzZzoR commented on 2014-01-14 18:11

...or you can open PKGBUILD, find md5sums=(...) and change first one (!) to this MD5 value - D37854CA30D75E3DE2C8D6C213B6ADC7

Actual for Chrome 33.0.1750.29 dev.

Det commented on 2014-01-14 09:47

Yes, there's a new one up. Use "--skipinteg" and pkgver=33.0.1750.27, if you don't t wanna wait.

Det commented on 2014-01-14 09:47

Yes, there's a new one up. Use "--skipinteg" and pkgver=33.0.1750.27, if you don't t wanna wait.

lockheed commented on 2014-01-14 08:19

The checksum fails for the main package and so prevents installation.

Det commented on 2014-01-08 19:44

HAPPY??

Det commented on 2014-01-08 15:33

Because I'm a fool!

danyf90 commented on 2014-01-08 13:20

why is the line "# [ "$CARCH" = 'x86_64' ] && _arch=amd64" commented out?

on 64 bit system it downloads the 32 bit version and the checksum fails

jnbek commented on 2013-12-27 04:56

Wow Det, removing all the files in /etc/ is hard to recover from >_< nice unix-fu!!

colinkeenan commented on 2013-12-26 22:34

Thanks Det,

I deleted libgcrypt.so.11 and what it linked to and re-installed google-chrome-dev. Works!

Det commented on 2013-12-26 13:45

The gcrypt library is now included and is installed even, if the current libgcrypt version is less than 1.6.0-1. This allows for getting the previous libgcrypt.so _before_ upgrading to the new one.

It took me a little while to say this, because I managed to remove all of my files in /etc.

Det commented on 2013-12-26 11:13

There have not been any known cases of successful disassembles of the closed-source Google Chrome internet application so far. Indisputably it was provided us by God.

Also, instead of downloading the library from a separate location, you can also do a: "pacman -S core/libgcrypt" to download the previous version from the official Arch Linux [core] repository.

Anyway, what we can do is to simply have the correct libgcrypt.so.11 installed to /opt/google/chrome(-unstable)/ for now, in case libcgrypt-1.6+ is installed on the machine.

colinkeenan commented on 2013-12-26 06:55

arzeth, Thanks.

I don't have the older version in /var/cache... So, I used your dropbox version. I guess that might be better than linking to the Steam version I was already using - at least it's slightly newer.

Does anyone know if it's possible to build google-chrome-dev so that it uses libgcrypt.so.20 instead though?

arzeth commented on 2013-12-26 02:55

colinkeenan, take this libgcrypt.so.11.8.2 from previous libgcrypt:
wget https://www.dropbox.com/s/02eiwajzbwofs9c/libgcrypt.so.11.8.2 -O /usr/lib/libgcrypt.so.11.8.2
ln -sT /usr/lib/libgcrypt.so.11.8.2 /usr/lib/libgcrypt.so.11

or...
pacman -U /var/cache/pacman/pkg/libgcrypt-1.5.3-1-x86_64.pkg.tar.xz
cp /usr/lib/libgcrypt.so.11.8.2 /tmp/
pacman -S libgcrypt
cp /tmp/libgcrypt.so.11.8.2 /usr/lib/
ln -sT /usr/lib/libgcrypt.so.11.8.2 /usr/lib/libgcrypt.so.11

colinkeenan commented on 2013-12-26 00:16

After today's updates (using testing), google-chrome won't start for me, complaining that it can't find libgcrypt.so.11. That's because I now only have libgcrypt.so.20. In order to get google-chrome to start, I did a symlink to the needed library that I happened to have in Steam:
sudo ln -s -T /home/colin/.local/share/Steam/ubuntu12_32/steam-runtime/amd64/lib/x86_64-linux-gnu/libgcrypt.so.11.7.0 /usr/lib/libgcrypt.so.11

Now google-chrome starts, but it complains that "The name org.freedesktop.secrets was not provided by any .service files". I don't know if that's related or not.

Anyway, I wanted to know if anyone could tell me how to make google-chrome-dev use /usr/lib/libgcrypt.so.20 instead of /usr/lib/libgcrypt.so.11?

Det commented on 2013-12-13 11:08

I know, but that's how they brain-deadedly provided them for us.

Anyway, I added a Chmod line.

Det commented on 2013-12-13 11:07

I know, but that's how they brain-deadedly provided them for us.

Anyway I added a chmod line.

danyf90 commented on 2013-12-13 10:57

yes but you have to change permission (e.g. 755) otherwise no one can access them

Det commented on 2013-12-13 10:48

It's just a leftover for compatibility reasons.

It only contains symlinks to /opt/google/chrome-unstable/.

danyf90 commented on 2013-12-13 10:22

what is the purpose of folder /otp/google/chrome/ ? it is accessible only by root

hexchain commented on 2013-12-11 02:54

The Dev Channel has been updated to 33.0.1734.0 for Linux.

ghaith commented on 2013-12-04 17:59

updated PKGBUILD here: http://pastebin.com/vCHGyYGS,
I had to comment out the .desktop entry as it was already in the appropriate location, also the /chrome links to /chrome-unstable as that's where I found them in the packages..

benklett commented on 2013-12-03 17:12

New release: 33.0.1726.0

sasy360 commented on 2013-11-22 08:24

Validating source files with md5sums...
google-chrome-unstable_33.0.1711.3_amd64.deb ... FAILED
eula_text.html ... Passed
Is the md5sum in PKGBUILD correct?

xvoprsalek commented on 2013-11-19 19:10

I liked the UI change.

phaseburn commented on 2013-11-19 18:57

The UI change has been reverted in 33.0.1711.3. Clipboard functionality, stability, sortable task manager columns, tab scroll wheel compatibility, etc, are all present again.

pritambaral commented on 2013-11-19 17:13

New upstream release 33.0.1711.3

jnbek commented on 2013-11-19 00:25

yea, I'm having problems with Copy/Paste, it seems to cause the process to crash and go into Z state. Closing and reopening the tab fails to correct the problem and even closing the entire browser doesn't help, one must SIGKILL the parent process to even get Chrome to start again... I sure hope they get this crap fixed because I rely on Chrome to do my work, and since I've never once had any problems with Unstable, until this release, I exclusively use Unstable everywhere. Unfortunately, I've upgraded all my work systems to 33 so I can't really work efficiently. Any word on when upstream intends on fixing these problems?

Det commented on 2013-11-17 09:25

This is why I love the Stable Channel.

xvoprsalek commented on 2013-11-16 19:24

Just maximize the window. lol
http://i.imgur.com/bGTcpuO.png

hendry commented on 2013-11-16 11:49

Chrome 33 can't seem to be able to communicate with my terminal's clipboard.

I've also highlighted in red the waste of space their new UI introduces:
http://s.natalian.org/2013-11-16/lost-space-33.png

jdarnold commented on 2013-11-13 14:14

And the flash player is out of date, so even Google Music doesn't work any more.

Sigh. Back to Firefox.

Kwpolska commented on 2013-11-13 13:50

WARNING: version 33 of Chrome introduces the following breakage:

a) system titlebars cannot be used anymore
b) one cannot use the scroll wheel anymore to switch tabs
c) …and many other things caused by an UI overhaul.

quarter82 commented on 2013-11-12 09:50

google-chrome-beta just hit the same version as google-chrome-dev

The problem occurs in both, however some sites work just fine. For example, *.qt-project.org/* (didnt try every page, but it seems to work).

8B3CE273 commented on 2013-11-12 06:57

using intel HD4000, crashes

wxg4net commented on 2013-11-12 00:03

using a nVidia card. it crashes.

jdarnold commented on 2013-11-11 14:44

My Chrome is working fine and I'm using a nVidia card with the binary drivers. Hope I didn't jinx myself...

tuftedocelot commented on 2013-11-11 13:59

Hm, that could be it. My laptop has Intel graphics and no problems here.

quarter82 commented on 2013-11-11 09:50

Does everyone with this problem have a nVidia graphics card?

I have a laptop with an ATI card, and it works there. Maybe it is not related because I'm not using Arch on that machine.

laputaprince commented on 2013-11-06 17:30

with
32.0.1685.0-1
32.0.1700.2
after rm all profile,settings and restart. it still crashes.

wxg4net commented on 2013-11-05 02:55

@quarter82 me too

quarter82 commented on 2013-11-04 09:56

When I try to access the Settings, the page crashes too.

Using flags to disable HW accelaration did not work. These are the flags I tested: google-chrome-unstable --disable-accelerated-video --disable-accelerated-plugins --disable-accelerated-compositing --disable-accelerated-video-decode --disable-accelerated-2d-canvas --disable-accelerated-overflow-scroll --disable-accelerated-video-decode --disable-gpu

colinkeenan commented on 2013-11-03 03:28

I'm not having any issues but that's probably because I was having issues many months ago and eventually figured out that if I turned off hardware acceleration, everything's fine.

I suggest you turn off hardware acceleration assuming you can get to the settings before it crashes.

jorgemota commented on 2013-10-30 22:27

tuftedocelot, Me, same problem, deleted profile, reinstalled, and now is in every page, new tab, settings, sign into google account... etc. :(

tuftedocelot commented on 2013-10-30 16:58

With 32.0.1685.0-1, is anyone getting tabs that instantly crash on startup?

beardedlinuxgeek commented on 2013-10-24 12:25

Newest version is 32.0.1678.0, but you can also just skip the md5check

md5sums=('SKIP'
'SKIP')
[ "$CARCH" = 'x86_64' ] && md5sums[0]='SKIP'

colinkeenan commented on 2013-10-17 23:32

Thanks shock_one. The md5sum provided by shock_one is for the EULA. When you edit the build, replace the 2nd md5sum in parentheses with the one provided by shock_one and the build will work.

shock_one commented on 2013-10-17 15:10

New md5sum: a3a49767c80e92cca63b7a33aadfb2af

corruptmemory commented on 2013-10-17 14:38

Updated EULA breaking install:

==> Validating source files with md5sums...
google-chrome-unstable_32.0.1671.3_amd64.deb ... Passed
eula_text.html ... FAILED

Det commented on 2013-10-13 12:10

Rebuilding could've also sufficed.

Kwpolska commented on 2013-10-13 08:35

PS. setting the path to “/opt/google/chrome-unstable/google-chrome” fixed the problem for me.

Kwpolska commented on 2013-10-13 08:26

Since the last update (chrome 32), the launcher is broken. It uses an executable path that the KDE icon-only task manager does not like, which results in me being unable to have one nice icon for chrome.

Det commented on 2013-10-11 15:51

That didn't last too long. Fixed.

palmtree5 commented on 2013-10-09 19:00

I had to edit and remove the "-unstable" as misc said. After that, it worked fine

misc commented on 2013-10-09 18:25

Remove all nine "-unstable" from the build() for this version.

oconnors11 commented on 2013-10-09 15:05

I am getting this any ideas?

-> Extracting google-chrome-unstable_32.0.1664.3_amd64.deb with bsdtar
==> Removing existing pkg/ directory...
==> Entering fakeroot environment...
==> Starting package()...
-> Extracting the data.tar.lzma
-> Moving stuff in place
mv: cannot stat ‘/var/build/google-chrome-dev/pkg/google-chrome-dev/opt/google/chrome-unstable/google-chrome-unstable.desktop’: No such file or directory
==> ERROR: A failure occurred in package().
Aborting...

Det commented on 2013-10-06 07:58

Well, don't ask here, ask in the forums (https://bbs.archlinux.org/viewforum.php?id=32).

kalpik commented on 2013-10-06 07:30

I'm running pulseaudio. I see a chrome stream on it when something is playing on chrome, but no sound output!

Det commented on 2013-10-06 07:26

I remember having this with bare ALSA I think.

Googling produces a zillion results: https://www.google.com/search?num=100&safe=off&q=site%3Aarchlinux.org+flash+no+sound

kalpik commented on 2013-10-06 07:15

I seem to have lost the ability to play sound on flash sites :(

Tried with both pepper flash, and adobe flash. Any ideas?

gnustomp commented on 2013-10-03 05:04

Det commented on 2013-09-29 15:54

You can always set your default browser for XDG with:

$ xdg-settings set default-web-browser google-chrome-unstable.desktop

Making (the Dev Channel) Chrome do this from its own button is a lot trickier than just changing the name of the symlink in /usr/bin and the .desktop.

Det commented on 2013-09-29 15:53

You can always set you default browser for xdg with:

$ xdg-settings set default-web-browser google-chrome-unstable.desktop

Making (the Dev Channel) Chrome do this from its own button is a lot trickier than just changing the name of the symlink in /usr/bin and the .desktop.

jnbek commented on 2013-09-24 02:26

can we please recreate the symlink to /usr/bin/google-chrome and update the .desktop files what not to function again? MATE/KDE/LXDE etc refuse to allow you to set Google Chrome as the default browser anymore. This is extremely annoying. I suppose it's due to the binary's name getting the -unstable appended to it, but there should be a good way to restore the Default Browser functionality. It used to show in Default Applications, but no longer does.

kremix commented on 2013-09-23 04:29

can't make it default browser in kde: make default clicked inside browser does nothing

Raymondcal commented on 2013-09-21 19:50

Ran the command (by myself, without rebuilding, you’re absolutely right Det!), and it works pretty well! Thanks again for your good work.

Det commented on 2013-09-21 09:32

Added 'gtk-update-icon-cache' in post_upgrade().

You can just run "gtk-update-icon-cache -q -t -f /usr/share/icons/hicolor" yourself though, so really there's once again absolutely no reason whatsoever to start nagging every single user to rebuild (or 'repackage', whatever you wanna call it) an entire package just to run one command.

LordAro commented on 2013-09-21 09:09

As far as I can tell, there is no longer an icon included (none in the /opt dir)

Of course, this could be a bug in chrome itself, rather than the PKGBUILD

Raymondcal commented on 2013-09-21 05:38

Hi!
Is there something wrong with the .desktop file? The icon is not the Chrome icon anymore; instead, I have the default Gnome executable icon. And in Gnome 3, I can’t set up Chrome-unstable as the default browser anymore.
Many thanks anyway for maintaining this aur file! :)

Det commented on 2013-09-20 19:19

Finally got it updated.

Good thing they did this. Certainly gonna give them more testers.

E: Yeah, and to clear out the '-dev conflict' tag from an already installed google-chrome(-beta) you need to either:

1) rebuild google-chrome(-beta) OR

To speed things up:

2) for google-chrome: change %CONFLICTS% in "/var/lib/pacman/local/google-chrome-<version>/desc" to 'google-chrome-beta' OR

3) for google-chrome-beta: simply remove the whole section, as google-chrome-dev has to provide 'google-chrome' so we can't put the conflict on both sides.

Det commented on 2013-09-20 18:52

Finally got it updated.

Good thing they did this. Certainly gonna give them more testers.

E: Yeah, and you either need to rebuild google-chrome(-beta) or alternatively change %CONFLICTS% in "/var/lib/pacman/local/google-chrome(-beta)-<version>/desc" to google-chrome(-beta) (the other one) to save time.

Det commented on 2013-09-20 18:49

Finally got it updated.

Good thing they did this. Certainly gonna give them more testers.

Det commented on 2013-09-20 18:28

http://pastebin.com/raw.php?i=02avqKjf

vorbote commented on 2013-09-20 11:52

Here is another (better) PKGBUILD. Icons need to be renamed to be found and the license directory should be renamed as well if google-chrome-unstable is to be installable alongside stable OR beta (just checked that myself).

https://gist.github.com/palopezv/6636337

And PLEASE, PLEASE people. Don't use pastebin.com Read <https://wiki.archlinux.org/index.php/List_of_Applications/Internet#Pastebin_clients> for the reasons why.

trizen commented on 2013-09-20 08:42

Here is a working PKGBUILD: http://pastebin.com/8cbcQZNd

Det commented on 2013-09-20 01:54

I can later today. It's a bit difficult to do that from the Finnish Army.

vorbote commented on 2013-09-20 01:32

Chromium.org people has finally seen the light and now google-chrome-unstable can be installed side by side with other channels. Now the directory is called google-unstable and the desktop and icon files are called google-chrome-unstable (from reading the desktop file) as well as the man page and the cron job file. Can you fix the PKGBUILD so it is installable?

I wonder if the beta channel PKGBUILD needs to be adjusted as well.

Det commented on 2013-09-07 08:21

Thanks once more. You really don't need to do that.

colinkeenan commented on 2013-09-06 07:53

There is a new version: 31.0.1622.0
64 bit .deb checksum: 7e26f7cf1ed00b7c6bf7b0b522e42d0a

Det commented on 2013-08-16 09:16

Thanks for the notifications I guess.. xD

colinkeenan commented on 2013-08-16 04:35

There is a new version: 30.0.1599.10
64 bit .deb checksum: 97f21038dd02258ae23e2e41ac249111

colinkeenan commented on 2013-08-13 16:05

There is a new version: 30.0.1599.0
64 bit .deb checksum: 014731b4af43217e5b41a4cb6480af88

colinkeenan commented on 2013-08-07 22:58

There's a new version: 30.0.1588.0

checksum: fc5cde1f113b44d5994a5d62cba946db for the 64 bit .deb file

colinkeenan commented on 2013-08-07 15:56

There's a new version: 30.0.1588.0.1-1 (can't use dashes though for version when editing the install file, so just 30.0.1588.0)

checksum: fc5cde1f113b44d5994a5d62cba946db for the 64 bit .deb file

Anonymous comment on 2013-08-01 16:54

Got it working now.
I had to modify the MD5 as per the downloaded deb.

In case anyone needs it, here's the working PKGBUILD: http://ix.io/700

dlsolo commented on 2013-07-30 23:43

I got it working....somehow. Deleted everything and re-downloaded the tarball. used this to build the package:

$ makepkg -s --skipchecksums --skipinteg

Package installed without issue.

dlsolo commented on 2013-07-30 23:34

Edit to my last comment:

When I add the e589... md5 into the PKGBUILD file, it passes the validity check, but fails due to:

==> Starting package()...
-> Extracting the data.tar.lzma
./opt/google/chrome/chrome.pak: Lzma library error: Corrupted input data
bsdtar: Error exit delayed from previous errors.
==> ERROR: A failure occurred in package().
Aborting...

dlsolo commented on 2013-07-30 23:32

@codekoala, thanks for the md5. I dropped that into that appropriate line, but the "one or more files did not pass the validity check!" remains. The version of google-chrome-unstable I have in my folder (to include src folder) is google-chrome-unstable_30.0.1573.2_amd64.deb. The md5 for that file is e589fd350cf499ce3cff5c53a75686d6. If I plug that md5 in, no dice either. Thoughts?

codekoala commented on 2013-07-30 18:08

This line:

[ "$CARCH" = 'x86_64' ] && md5sums[0]='af1e95971b40ea087d3db999ce5a7c9c'

codekoala commented on 2013-07-30 18:07

@jdarnold, @dlsolo - you might need to modify the line where the hash is set for the 64-bit .deb. By default, updpkgsum will update the hash for the 32-bit .deb.

codekoala commented on 2013-07-30 18:07

@jdarnold, @dlsolo - you might need to modify the line where the hash is set for the 64-bit .deb. By default, updpkgsum will update the hash for the 32-bit .deb.

jdarnold commented on 2013-07-30 18:05

yeah, same here. Been meaning to ask about it.

dlsolo commented on 2013-07-30 17:51

I'm getting an error due to md5 mismatch. i've tried makepkg -g, updpkdsum, and md5sum on both the eula and .deb files, with no luck. Is there another way of installed this variant of Google Chrome? Thanks in advance.

Det commented on 2013-07-19 15:49

@acl, I can, but I'd still need to bump the pkgver whenever there's a new one up, since not everybody bothers to do that.

pierrec commented on 2013-07-19 12:52

pkgver=30.0.1568.0
pkgrel=1

md5sums=('50d64cb826de72e9e630984b20a19ac0'
'6d57da7476a4b1b7a81821d9c036425c')
[ "$CARCH" = 'x86_64' ] && md5sums[0]='4160ca94c0adb233fb9b55f8d6eb1a8a'

pierrec commented on 2013-07-19 12:48

pkgver=30.0.1568.0
pkgrel=1

md5sums=('50d64cb826de72e9e630984b20a19ac0'
'6d57da7476a4b1b7a81821d9c036425c')
[ "$CARCH" = 'x86_64' ] && md5sums[0]='4160ca94c0adb233fb9b55f8d6eb1a8a'

pierrec commented on 2013-07-19 12:45

pkgver=30.0.1568.0
pkgrel=1

md5sums=(''
'6d57da7476a4b1b7a81821d9c036425c')
[ "$CARCH" = 'x86_64' ] && md5sums[0]='4160ca94c0adb233fb9b55f8d6eb1a8a'

Anonymous comment on 2013-07-17 11:55

I'm using Yaourt to get the package, and can't install it due to the signature thing, is there an arg to filter that on Yaourt? if not, you might wanna get look into that.

Also since this is a dev meta, can't you refer to the latest package without hardcoded version strings? Just guessing here probably I'm 100% wrong, just leaving it out there anyway.

Thanks for the effort, tho :)

spsf64 commented on 2013-07-17 02:21

Ver 30.x is out!

codekoala commented on 2013-06-27 13:25

@Det: Touche :)

I was just asking because it was added some road bumps in my normal workflow when maintaining my own repository of packages like this. Thanks for the clarification!

Det commented on 2013-06-26 19:57

1) Yes.

2) Well, using 'updpkgsums && makepkg -si' instead of 'makepkg -si --skipinteg' is only worthwhile if you're building multiple times in a row - which you wouldn't be with this.

updpkgsums is meant as a replacement for 'makepkg -g >> PKGBUILD'. You don't need to use it, if you're not uploading anything.

codekoala commented on 2013-06-26 13:59

@Det I'm a bit confused by the PKGEXT change. What's the point in doing that? Simply to speed up build/install time by removing the compression step?

Also, there's a new utility to update the hashes automatically: updpkgsums. Run that in the same directory as a PKGBUILD and it will make sure the sums are correct, so you shouldn't need to use --skipinteg.

Det commented on 2013-06-25 19:40

So I changed the source urls to .debs (up to 12M less download, cutting off roughly ~21%, along with faster extraction times), added the license and changed the compression to PKGEXT=".pkg.tar".

The urls now also always point to the latest release, so in case I'm not there in time all you need to do is check the pkgver with either:

1) http://googlechromereleases.blogspot.com/search/label/Dev%20updates

2) curl -s https://dl.google.com/linux/direct/google-chrome-unstable_current_x86_64.rpm | head -c96 | strings

3) curl -s http://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | gzip -df | awk -F\" '/pkgid/{ sub(".*-","",$4); print $4": "$10 }'

and then rebuild with '--skipinteg'.

t3ddy commented on 2013-06-25 18:00

Perfect!

Det commented on 2013-06-25 17:57

No, I got my own. Thanks.

t3ddy commented on 2013-06-25 17:57

Good, do you want the sripts I was using?

Det commented on 2013-06-25 17:55

Owned

t3ddy commented on 2013-06-25 17:55

Disowned

Det commented on 2013-06-25 17:52

memememememememememememememe

t3ddy commented on 2013-06-25 17:40

Sorry guys, I've been busy in the real world and... my arch box has gone :(
Anyone wants to maintain chrome's pkguild?

colinkeenan commented on 2013-06-19 16:05

Thanks Det,

Here's what I ended up doing:

1) Downloaded google-chrome-unstable_current_x86_64.rpm from http://dev.chromium.org/getting-involved/dev-channel under the heading of 64-bit Fedora/Red Hat/OpenSUSE.

2) md5sum *.rpm (in the Downloads directory, of course)

3) When yaourt asked if I wanted to edit PKGBUILD, I said yes and both corrected the version in 2 places and corrected the 64 bit md5sum.

For 64 bit version, the md5sum is currently 57e61d610cafd4abbeb28cf7d226e42f

Det commented on 2013-06-19 08:38

You don't need to do _anything_. You can either:

1) Generate the md5sums with "updpkgsums" (since pacman 4.1),

2) manually put it in place by doing something like "md5sum *.rpm" and editing the PKGBUILD yourself OR

3) just build with "makepkg -si --skipinteg", since we couldn't care less about the md5sum we can't verify with anything else anyway.

colinkeenan commented on 2013-06-19 07:02

curl -s https://dl.google.com/linux/direct/google-chrome-unstable_current_x86_64.rpm | head -c96 | strings

returns the current version, and it is

google-chrome-unstable-29.0.1541.2-207000

So, I can edit PKGBUILD and fix the version in 2 places which allows the source code to be downloaded

Now, what do I do about the md5 checksum so it can be verified and built?

dr-peppa commented on 2013-06-12 04:12

The source URL does not work. (i.e. Error 404, not found)

dp

dr-peppa commented on 2013-06-12 03:48

The source URL does not work. (i.e. Error 404, not found)

dp

leitecarvalho commented on 2013-06-08 19:12

@det isn't troublesome to update this package every week? i've started to see some packages here that are built from git and skip the md5sum. maybe it would be useful in this package: md5sum('SKIP')

@tsuujin i've changed to "use GTK+ Theme" and the lines are gone.

Anonymous comment on 2013-06-08 15:19

I'm having an issue with this build's location bar icons. See the link for a screenshot of the issue. The "locked" and "star" icons have lines through the background, which seems to happen with all themes except for the gtk one.
http://minus.com/lbp3RS4Im3iePy

Actually, on further review it looks like the entire location bar has the lines through it, and the text just has an opaque white background so you can't see it.

Det commented on 2013-06-07 17:13

Uuhhh, yeah, you could do that locally anwyay, but the way we do things here in the Arch world is by using the _actual_ version numbers so that people only get to download this thing when it's required.

I'm not sure what else would you mean by that.

leitecarvalho commented on 2013-06-07 16:58

afterthought: on gentoo, we had the -9999 package version for latest versions of the package. i've read the Arch pkg guidelines but didn't find any relevant info.

leitecarvalho commented on 2013-06-07 16:37

https://dl.google.com/linux/direct/google-chrome-unstable_current_x86_64.rpm

md5: 180ca5354113cc0196146634a986bfb0

i did the md5sum as i couldn't find where google places its md5. Also, note the change of protocol ( https), the url and the use of "current" instead of package version.



pkgver='current' # Check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Dev%20updates

[...]

_verbld='current'
[...]
md5sums=('180ca5354113cc0196146634a986bfb0')
[...]
source=("https://dl.google.com/linux/direct/google-chrome-${_channel}_${_verbld}_${_arch}.rpm")

Det commented on 2013-05-15 13:41

The new one's ttf-google-fonts-git (https://aur.archlinux.org/packages/ttf-google-fonts-git/?comments=all):

"Comment by jstjohn
2013-04-25 03:08
Google Web Fonts has been rebranded as Google Fonts[1]. [...]

[1] http://9to5google.com/2013/04/22/google-web-fonts-rebrands-as-google-fonts-simplifies-web-interface/
[...]"

mbunkus commented on 2013-05-15 12:28

Optional dependencies include the package ttf-google-webfonts which does not exist according to "yaourt -S ..." and "yaourt -Ss ttf google".

Anonymous comment on 2013-03-28 19:21

For easier updating, one can separate out the SVN revision number as in http://pastebin.com/c0pNQdDs.

The new values are _svnver=191032 and pkgver=27.0.1453.6

Ragnis commented on 2013-03-24 14:57

New PKGBUILD for 27.0.1448.0:

http://pastie.org/7099920

scrtyfrk commented on 2013-03-06 06:22

Hello... Here's the PKGBUILD for the build 27.0.1430.0-186115

http://pastebin.com/CsY8bJa4

Anonymous comment on 2013-03-06 03:54

download of the 64bit rpm fails

jnbek commented on 2013-03-04 17:51

@ Wilberfan, I know right?

wilberfan commented on 2013-03-03 00:34

Boy, as a complete noob to applying patches, It'd sure be helpful to see someone elses (working) PKGBUILD... I've been studying the Wiki...but a working reference file would make an excellent study guide!

:-D

Anonymous comment on 2013-03-02 07:28

Here is a quick patch for your PKGBUILD for the Problem described from @tuftedocelot, @riccierim, @rickeyski and me ;-)

https://gist.github.com/jk779/5070051

Anonymous comment on 2013-03-02 06:54

I'm just curious, is this dev-Version of Chrome as safe as the stable, concernining security?
Does -dev get the same security-patches than -stable for example?
Anything else to mention? (Except that a dev version is not bug-free ;-)

Det commented on 2013-03-01 11:13

"Shameless" :D. Are you kidding me?

E: Oh yeah, you meant as in the fix you borrowed from them was "stolen" but I don't understand why would you even want to do that when the current systemd libudev.so works just fine.

Det commented on 2013-03-01 07:32

"Shameless" :D. Are you kidding me?

rickeyski commented on 2013-03-01 00:14

here is a pkgbuild fix for the strip option as well as the shameless borrowing of gentoo's libudev fix https://gist.github.com/rickeyvisinski-kanban/5061325

HarD commented on 2013-02-28 05:44

I have error

/usr/lib/yaourt/pkgbuild.sh: line 200: 19339 PKGDEST="$YPKGDEST" makepkg "${MAKEPKG_ARG[@]}" -s -f -p ./PKGBUILD

riccieri commented on 2013-02-28 02:51

@tuftedocelot I am. I solved this by negating "strip" from the OPTIONS array on /etc/makepkg.conf

tuftedocelot commented on 2013-02-28 02:23

Anybody receive this error message when updating?: http://bpaste.net/raw/80368/

Det commented on 2013-02-27 15:20

Well, that's why it's out of date..

evanlec commented on 2013-02-27 14:51

latest source link doesn't work:

http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-unstable-26.0.1410.12-183726.x86_64.rpm

404..

so cannot build package now

evanlec commented on 2013-02-27 14:51

latest source link doesn't work:

http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-unstable-26.0.1410.12-183726.x86_64.rpm

404..

so cannot build package now

HarD commented on 2013-02-27 05:27

Does print dialog is working on google-chrome-dev x86_64? I cann't print from it, only from system print dialog...

Det commented on 2013-02-05 12:04

Because then you don't have to do a:

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

...to bypass the previous tarball with the same name (which I would, but it's not my call).

c4software commented on 2013-02-05 11:40

I'm curious, why are you not using this two links in the package?

https://dl.google.com/dl/linux/direct/google-chrome-unstable_current_i386.rpm
and
https://dl.google.com/dl/linux/direct/google-chrome-unstable_current_x86_64.rpm

t3ddy commented on 2013-01-17 20:06

Thanks Det! I missed it :)
It will be fixed in the next version

Det commented on 2013-01-17 19:18

t3ddy, I suppose you didn't see the comment about 'openssl098'. It's obsolete now:

└┌(%:~/Desktop)┌- ldd /opt/google/chrome/PepperFlash/libpepflashplayer.so | grep libcrypto.so.0.9.8
┌┌(det@Archlinux)┌(416/pts/1)┌(09:14pm:01/17/13)┌-

Det commented on 2013-01-11 12:16

Btw. the Pepper Flash plugin no longer links to "libcrypto.so.0.9.8" so 'openssl098' is no longer needed.

captainju commented on 2013-01-11 11:38

@martincanaval
I have the same problem, so i switched to beta channel

martincanaval commented on 2013-01-11 03:31

Couple things:

1. I've been using the beta and dev versions with pepper flash without openssl098 and no issues, so not sure if that is still needed. (maybe some feature I haven't used)

2. I have the feeling that I'm the only person getting "segmentation fault" errors with the dev version (because I haven't found info about it). It's been happening to me for a past 5 or so versions. Beta and Stable versions work like a charm. I'll try again when this version gets updated.

Det commented on 2012-12-19 13:13

LOL. Accident..

captainju commented on 2012-12-14 07:32

error 404
A new version is available
http://googlechromereleases.blogspot.fr/2012/12/dev-channel-update_13.html

Det commented on 2012-12-11 16:00

Mh, what that command does (without the ";" or "&&" in between) is that it tries to remove the non-existing "rm" in addition to the actual files (you're also specifying '-f' twice).

So it's either this:
rm -f "$pkgdir/opt/google/chrome/nacl_irt_x86_64.nexe" "$pkgdir/opt/google/chrome/nacl_irt_srpc_x86_64.nexe"

..or in two lines as t3ddy finally did there.

mamamia88 commented on 2012-12-11 14:22

Don't ask me why but line 45 has to read all one line otherwise it doesn't work

rm -f "$pkgdir/opt/google/chrome/nacl_irt_x86_64.nexe" rm -f "$pkgdir/opt/google/chrome/nacl_irt_srpc_x86_64.nexe"

Det commented on 2012-12-08 19:37

Ooh, it's some "srpc" binary. Yeah. Well. He should do that. Maybe even some asterisk (*) setup to deal with those in the feature versions too.

Flagging.

bb010g commented on 2012-12-08 19:08

I've still got the same problem as ninian (with same error). Fails in both aura and manual building.

Det commented on 2012-12-07 22:35

@ninian, this was fixed back when you posted the same thing for the second time.

ninian commented on 2012-12-07 19:53

Adding this after line 45 in the PKGBUILD seems to do the trick (on my 32-bit system):
rm -f "$pkgdir/opt/google/chrome/nacl_irt_srpc_x86_64.nexe"

ninian commented on 2012-12-07 19:39

Latest version fails to build for me (using aurget -Sb), reporting the error:
-> Stripping unneeded symbols from binaries and libraries...
strip:./opt/google/chrome/nacl_irt_srpc_x86_64.nexe: File format not recognized
/usr/bin/fakeroot: line 181: 1349 User defined signal 1 FAKEROOTKEY=$FAKEROOTKEY LD_LIBRARY_PATH="$PATHS" LD_PRELOAD="$LIB" "$@"

ninian commented on 2012-12-07 19:23

Latest version fails to build for me (using aurget -Sb), reporting the error:
-> Stripping unneeded symbols from binaries and libraries...
strip:./opt/google/chrome/nacl_irt_srpc_x86_64.nexe: File format not recognized
/usr/bin/fakeroot: line 181: 1349 User defined signal 1 FAKEROOTKEY=$FAKEROOTKEY LD_LIBRARY_PATH="$PATHS" LD_PRELOAD="$LIB" "$@"

ninian commented on 2012-12-07 18:24

Latest version fails to build for me (using aurget -Sb), reporting the error:
-> Stripping unneeded symbols from binaries and libraries...
strip:./opt/google/chrome/nacl_irt_srpc_x86_64.nexe: File format not recognized
/usr/bin/fakeroot: line 181: 1349 User defined signal 1 FAKEROOTKEY=$FAKEROOTKEY LD_LIBRARY_PATH="$PATHS" LD_PRELOAD="$LIB" "$@"

wolftankk commented on 2012-12-07 06:47

25.0.1349.2
RPM download url:
http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-unstable-25.0.1349.2-171264.x86_64.rpm

wolftankk commented on 2012-12-07 06:44

25.0.1349.2
RPM download url:
http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-unstable-25.0.1349.2-171264.x86_64.rpm

Anonymous comment on 2012-12-07 05:27

25.0.1349.2 for Linux is out.

prash commented on 2012-11-30 16:10

I was wrong to have flagged it out of date. Apparently, the latest (25.0.1337.0) is available only for Windows, Mac, and Chrome Frame.

Anonymous comment on 2012-11-25 14:17

it gets hang every now and then, while playing flash videos. after few minutes it shows the message shockwave not responding kill it or wait .

t3ddy commented on 2012-11-24 09:19

I can download it without any problem

Anonymous comment on 2012-11-24 07:01

I think there is something wrong with the link :https://aur.archlinux.org/packages/go/google-chrome-dev/google-chrome-dev.tar.gz

I can't download the right file "google-chrome-dev.tar.gz"。。。

Hspak commented on 2012-11-23 23:57

Does anyone get a 'Media Cache' folder in your home directory from a recent update? It still has the usual cache folder under ~/.cache, but I noticed this folder getting created just now.

Det commented on 2012-11-20 18:34

WRONG! >:O

kureshii commented on 2012-11-19 11:19

After upgrading icu to 50.1, I get the following error launching google-chrome:

/usr/bin/google-chrome: error while loading shared libraries: libicuuc.so.49: cannot open shared object file: No such file or directory

Seems icu-49.* is an explicit dependency and should be added in depends. Not sure if future versions of chrome will update this dependency but I remember running into it a few months ago after another icu update as well.

Det commented on 2012-11-18 23:21

It is a stupid one. If you wanna use the 24. series use the Beta Channel.

printesoi commented on 2012-11-18 23:07

Excuse for the stupid question but where I can find the rpm of the previous version? (24.0.1297.0-162078)

http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-unstable-24.0.1297.0-162078.x86_64.rpm

If I try the link above, it gives me 404 Not Found..

lockheed commented on 2012-11-14 12:09

Thanks. I am well aware of that unwise decision by google and the general outrage that followed it.

Det commented on 2012-11-14 12:06

Then you're in trouble. It was deprecated a long time ago and I couldn't find any such extensions in the Chrome Web Store either.

E: In fact, see: http://code.google.com/p/chromium/issues/detail?id=85184#c26

Det commented on 2012-11-14 12:02

Then you're in trouble. It was deprecated a long time ago and I couldn't find any such extensions in the Chrome Web Store either.

lockheed commented on 2012-11-14 11:52

I need it to be hidden in non-maximized windows.

Det commented on 2012-11-14 11:50

It's called the omnibox and what's wrong with F11?

lockheed commented on 2012-11-14 11:42

Is Compact Navigation feature available in dev channel? If not, is there any other way to autohide the URL and TAB bars?

wolftankk commented on 2012-11-13 06:52

google-chrome-dev has updated 25.0.1323.1

download url:
http://dl.google.com/linux/chrome/rpm/stable/x86_64/google-chrome-unstable-25.0.1323.1-167142.x86_64.rpm

ravicious commented on 2012-11-06 19:45

I get the same error as MajorTom:

curl: (22) The requested URL returned error: 404 Not Found

Anonymous comment on 2012-10-31 05:11

current version should be 24.0.1312.1-164922 judging by the wget stuff

WiZeTeK commented on 2012-10-17 02:21

curl: (22) The requested URL returned error: 404 Not Found
==> ERROR: Failure while downloading google-chrome-unstable-24.0.1290.1-160787.i386.rpm

dcelasun commented on 2012-10-15 16:18

For me, Chrome 24 broke GPU acceleration with Catalyst (worked fine with 23). "about:gpu" shows "GpuProcessHostUIShim: GPU Process Crashed." and the GL information is completely empty. Screenshot: http://i.imgur.com/DGLoT.png

ruario commented on 2012-08-27 07:58

@Ner0: Pulling the information from the header of the rpm package or by querying the YUM or APT databases is better than relying on pkgs.org to be up to date. After all Google are the ones updating the packages, so it is best to get the information directly from them rather than a third party.

Det commented on 2012-08-24 07:52

Wasn't just now.

Ner0 commented on 2012-08-15 13:25

You could also check http://pkgs.org/google-chrome-for-linux they're always up to date

ruario commented on 2012-08-15 09:53

@jdarnold: You can also work out the build version (including minor revision) numbers by reading it from the first 96 bytes of the of the rpm header, e.g.:

$ curl -sr0-96 http://dl.google.com/linux/direct/google-chrome-unstable_current_i386.rpm | strings

sjakub commented on 2012-08-15 07:57

22.0.1229.6-151414 is there

riccieri commented on 2012-08-09 21:21

I've updated it to 22.0.1229.2

full file: http://pastebin.com/fhYWrDsc
patch from 22.0.1221.1: http://pastebin.com/Pkp3nm06

t3ddy commented on 2012-08-02 11:03

I use a script to update chrome, here the commands to find out the version
wget -qO- http://dl.google.com/linux/chrome/rpm/stable/i386/repodata/other.xml.gz | zcat | grep "google-chrome-${CHAN}" | awk -F\" '{print $10"-"$12}'
wget -qO- http://dl.google.com/linux/chrome/rpm/stable/x86_64/repodata/other.xml.gz | zcat | grep "google-chrome-${CHAN}" | awk -F\" '{print $10"-"$12}'

jdarnold commented on 2012-08-01 11:54

Thanks for updating the PKGBUILD. But I'm curious - how do you find out the build number? I went on a quest last night to figure out the actual file to download but couldn't find out the info, esp the -149058 part.

Anonymous comment on 2012-08-01 07:17

Your link says: "(Update: also 22.0.1221.1 for Linux)" :P

Det commented on 2012-07-31 09:19

Just for Windows and Mac this one, not out-of-date: http://googlechromereleases.blogspot.fi/2012/07/dev-channel-update_30.html

pawelj84 commented on 2012-07-29 13:05

There is an update to 22.0.1215.0, correct lines for i386:
pkgver=22.0.1215.0
_verbld=22.0.1215.0-147830
md5sums=('ab0f40dad2eed62a27325a812a13268d')

Anonymous comment on 2012-07-26 12:03

OK, next time I'll pastebin the whole PKGBUILD.

Det commented on 2012-07-26 10:11

Yeah, we're all very grateful but this wonderful invention of pastebin is still left with sadness and suicidal thoughts in its lonely life because you guys aren't using it.

Anonymous comment on 2012-07-25 23:19

This is the updated header for the new version 22.0.1215.0-147830


pkgname=google-chrome-dev
pkgver=22.0.1215.0 # Check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Dev%20updates
pkgrel=1
pkgdesc="An attempt at creating a safer, faster, and more stable browser (Dev Channel)"
arch=('i686' 'x86_64')
url="http://www.google.com/chrome"
license=('custom:chrome')
depends=('alsa-lib' 'gconf' 'gtk2' 'hicolor-icon-theme' 'libpng' 'libxslt' 'libxss' 'nss' 'ttf-dejavu' 'xdg-utils')
optdepends=('kdebase-kdialog: needed for file dialogs in KDE' 'openssl098: needed for built-in flash-plugin to work')
provides=("google-chrome=$pkgver")
conflicts=('google-chrome')
install=${pkgname}.install
_channel='unstable'
_verbld=22.0.1215.0-147830

if [ "$CARCH" = "i686" ]; then
_arch='i386'
md5sums=('ab0f40dad2eed62a27325a812a13268d')
elif [ "$CARCH" = "x86_64" ]; then
_arch='x86_64'
md5sums=('3f080e3312a9e2a63322b6f2fcbf4669')
fi

ciferkey commented on 2012-07-25 19:33

Please update to reflect the new version 22.0.1215.0-147830.

Anonymous comment on 2012-07-10 12:05

Here is the header of the PKGBUILD with the new values:

pkgname=google-chrome-dev
pkgver=22.0.1201.0
pkgrel=1
pkgdesc="An attempt at creating a safer, faster, and more stable browser (Dev Channel)"
arch=('i686' 'x86_64')
url="http://www.google.com/chrome"
license=('custom:chrome')
depends=('alsa-lib' 'gconf' 'gtk2' 'hicolor-icon-theme' 'libpng' 'libxslt' 'libxss' 'nss' 'ttf-dejavu' 'xdg-utils')
optdepends=('kdebase-kdialog: needed for file dialogs in KDE' 'openssl098: needed for built-in flash-plugin to work')
provides=("google-chrome=$pkgver")
conflicts=('google-chrome')
install=${pkgname}.install
_channel='unstable'
_verbld=22.0.1201.0-145644

if [ "$CARCH" = "i686" ]; then
_arch='i386'
md5sums=('41a231f9578c7893651d4d231c79a9ca')
elif [ "$CARCH" = "x86_64" ]; then
_arch='x86_64'
md5sums=('70e9ed0d7cefba19579b6103e6f06a59')
fi

Det commented on 2012-06-29 07:30

It means there's a new version.

Anonymous comment on 2012-06-29 03:06

I'm confused. "ERROR: Failure while downloading google-chrome-unstable-21.0.1180.11-143993.x86_64.rpm"

Det commented on 2012-06-22 09:22

To be honest, I didn't test it neither, I just know it works.

t3ddy commented on 2012-06-20 17:10

Because I don't have a 32bit machine, so I can't test it :)
Updated

Det commented on 2012-06-20 16:55

Oh yeah, it does. But you probably want to use the '-f' flag, in case you're too lazy to see in which update will it be removed.

But so why didn't you update the package itself?

t3ddy commented on 2012-06-20 14:54

The pkbguild I've posted previously doesn't work?

Det commented on 2012-06-20 14:28

[ "$CARCH" == 'i686' ] && would do.

xzy3186 commented on 2012-06-20 13:33

Is it possible to add something like "if [ $CARCH == 'i686' ]; then ......" in your PKGBUILD?

t3ddy commented on 2012-06-20 11:51

That would be good for 32bit, but what about 64bit? I don't think removing its binary is a good move.

xzy3186 commented on 2012-06-20 11:21

The same error for 32bit. An easy way to get this around is adding "rm ./opt/google/chrome/nacl_irt_x86_64.nexe" to the first line of package() in PKGBUILD.

t3ddy commented on 2012-06-20 10:55

Since I don't have a 32-bit machine, can you try this: http://pastebin.com/jVe89qy2
Thanks

Anonymous comment on 2012-06-20 10:47

Same problem with new version.

t3ddy commented on 2012-06-16 17:35

For some (strange) reason there's a 64-bit executable into 32-bit rpm.
Let's see if in the next release it'll be still present, in that case I'll remove it using the pkgbuild.

t3ddy commented on 2012-06-16 17:29

This is not happening here. Are you on 32 bit?

Anonymous comment on 2012-06-16 11:23

Same error here...
strip:./opt/google/chrome/nacl_irt_x86_64.nexe: File format not recognized
/usr/bin/fakeroot: line 181: 19557 Segnale 1 definito dall'utente FAKEROOTKEY=$FAKEROOTKEY LD_LIBRARY_PATH="$PATHS" LD_PRELOAD="$LIB" "$@"

Is there a workaround?

Anonymous comment on 2012-06-13 05:45

-> Stripping unneeded symbols from binaries and libraries...
strip:./opt/google/chrome/nacl_irt_x86_64.nexe: File format not recognized
/usr/bin/fakeroot: line 181: 13402 User defined signal 1 FAKEROOTKEY=$FAKEROOTKEY LD_LIBRARY_PATH="$PATHS" LD_PRELOAD="$LIB" "$@"
==> ERROR: Makepkg was unable to build google-chrome-dev.

Anonymous comment on 2012-06-13 01:21

pacman -Q | grep yaourt
yaourt 1.1-1

==> ERROR: An unknown error has occurred. Exiting...
/usr/lib/yaourt/pkgbuild.sh: line 200: 2757 User defined signal 1 PKGDEST="$YPKGDEST" makepkg "${MAKEPKG_ARG[@]}" -s -f -p ./PKGBUILD
==> ERROR: Makepkg was unable to build google-chrome-dev.

t3ddy commented on 2012-06-02 14:08

I've added the workaround to the pkgbuild, thanks rememberthemer.

Anonymous comment on 2012-06-02 13:05

A better solution to the udev issue is to put the link in /opt/google/chrome.
ln -sf /usr/lib/libudev.so.1.0.1 /opt/google/chrome/libudev.so.0
Less chance of spurious junk being left behind + no chance of programs being built with an rpath to the wrong libudev soname.
Even better - add this to the package please.

Anonymous comment on 2012-06-01 15:17

solved:
ln -sf /usr/lib/libudev.so.1.0.1 /usr/lib/libudev.so.0

Anonymous comment on 2012-06-01 07:02

$ google-chrome
/usr/bin/google-chrome: error while loading shared libraries: libudev.so.0: cannot open shared object file: No such file or directory
$ locate libudev.so
/usr/lib/libudev.so
/usr/lib/libudev.so.0
/usr/lib/libudev.so.0.13.1

felixonmars commented on 2012-05-31 02:28

\\n: Name or service not known
Cannot handle "host" cmdline arg `\\n' on position 1 (argc 3)
\\n: Name or service not known
Cannot handle "host" cmdline arg `\\n' on position 1 (argc 3)
Unable to read PKGBUILD

mango commented on 2012-05-28 08:50

2 yangtsesu:
ln -s /usr/lib/libudev.so.1.0.0 /usr/lib/libudev.so.0

yangtsesu commented on 2012-05-28 01:40

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

oliva commented on 2012-05-24 19:10

Also, less often, but the same happens when opening a link in new tab (background by default).

magicfish1990 commented on 2012-05-24 14:04

I submitted an issue.
https://code.google.com/p/chromium/issues/detail?id=129561

And chromium-dev have same problem. https://aur.archlinux.org/packages.php?ID=37244

oliva commented on 2012-05-24 12:25

Same here, but with linux 3.4.0-1-ARCH (x86_64)

CasualSuperman commented on 2012-05-23 22:35

Same problem here.

magicfish1990 commented on 2012-05-23 05:53

Sometimes close the tab will cause Segmentation fault.
Anyone have similar problems?

Chrome Version 21.0.1145.0 dev.
Linux version 3.3.6-1-ARCH x86_64
glibc 2.15-10

jahiy commented on 2012-05-18 12:00

chrome-dev has updated to 20.0.1132.11-137611

jahiy commented on 2012-05-18 11:57

chrome-dev has updated to 20.0.1132.11-137611

Anonymous comment on 2012-05-06 08:23

now working for me in version 20.0.1123.4 dev

jnbek commented on 2012-04-29 01:48

I am having the exact same problem with this as tanguyr

Anonymous comment on 2012-04-26 20:14

Every time I try to access an encrypted Google site (gmail, reader, plus, etc), I get the following error:

Error 710 (net::ERR_KEY_GENERATION_FAILED): Unknown error.

Other (non Google) https sites don't have this problem.

This is with google-chrome-dev 20.0.1115.1-1 x86_64

zenixls2 commented on 2012-04-11 08:26

change to google-chrome-unstable-20.0.1096.1-131343.${_arch}.rpm ??
md5sum x86_64 8ae5e70e7d48280f61424169af67d5b6
md5sum i386 c5f994e9a7adeaccedb7f3d8ede6cbee
maybe _arch change to "current" is better

Det commented on 2012-04-10 06:50

You know, it's easier to just go to chrome:plugins and see if there's 2 flash plugins.

NobodySpecial commented on 2012-04-10 00:46

I'm running 32 bit - that must be the difference.

gumper commented on 2012-04-10 00:33

I could be wrong but my understanding is that it's currently not available for 64-bit versions of Chrome. Are you running 32 or 64 bit?
I have went back to using Chromium as I see no advantage to using Chrome if the built in flash is not available.

NobodySpecial commented on 2012-04-09 23:22

I thought this DID have built-in flash.
Here's what I get from http://flashbuilder.eu/flash-player-version.html
Firefox: 11.2.202.228
google-chrome-dev: 11.2.202.229

That leads me to think Firefox is flashplugin and google-chrome-dev is using built-in player.

t3ddy commented on 2012-04-09 15:59

Currently no.

gumper commented on 2012-04-09 13:23

Thanks for the reply. Is there any version of Chrome currently available for Linux that has this included with it? I also tried the package "google-chrome" from the AUR and it doesn't work for me there either.

t3ddy commented on 2012-04-09 09:14

Simply it's not included in this dev release.

gumper commented on 2012-04-08 03:14

Could someone tell me why I'm not seeing the built-in flash player when I go to "about:plugins"? The only one that I see there is the one that I have installed through pacman and if I disable that one, flash doesn't work at all.

I'm running version 19.0.1084.15 dev of chrome (64 bit) and I also have openssl098 version 0.9.8.u-1 installed.

thehoff commented on 2012-04-06 05:16

New dl link for chrome dev branch: http://dl.google.com/dl/linux/direct/google-chrome-unstable_current_x86_64.rpm.

Det commented on 2012-04-01 10:43

It's based on the same version. They update it with new releases.

Anonymous comment on 2012-04-01 00:48

Does somebody know why the build-in flash version is 11.2.31.118. The latest 'official' version is 11.2.202.228. Why doesn't it update automatically?

Det commented on 2012-03-29 22:40

By the way, this little 'documentation comic book' here is definitely a good read: http://www.google.com/googlebooks/chrome/

It may seem sort of beginner-intended at first but it really goes into great detail and gives a very good understanding of what's really going "down" under the hood when you're browsing with Chrome. It even brings up some earlier problems with browsers and explains why a completely new one was started.

And yeah, 39 pages may seem like a lot, but trust me that it's not: all that stuff is very "refreshingly" explained and (at least for me) it was pretty hard to wanna stop.

Anonymous comment on 2012-03-18 13:00

First letter of a url or search in omnibox does freeze the complete browser… not always, but again and again. :(

rickeyski commented on 2012-03-17 12:48

Weird, glitch in the matrix apparently, I upgraded one box and the next failed. Now everything is good again.

Det commented on 2012-03-16 20:01

I think the /usr/lib symlinks can be removed altogether (including the mozilla plugin one). Chrome doesn't require them anymore.

Also, if 'openssl098' was moved to optional dependencies this 'google-chrome-mini' clone could be removed: http://mailman.archlinux.org/pipermail/aur-general/2012-March/018154.html.

Anonymous comment on 2012-03-15 18:36

I have a problemn with the pepperflashplayer (libpepflashplayer.so)on different sides.
If I deactivate pepper flash and take the normal Flashplayer, then I do not have the problem

Det commented on 2012-03-15 17:34

By the way, for those still experiencing the random hangs: don't just clear the browsing data - create a new profile.

I removed ~/.cache/google-chrome/ and ~/.config/google-chrome/, synced everything back with Chrome's sync (thank god for this) and haven't had a single hang since.

Frankly, the whole thing even _feels_ so much faster now.

Det commented on 2012-03-15 16:32

It's .1068.1, not .1061: http://googlechromereleases.blogspot.com/2012/03/dev-channel-update_15.html

Anonymous comment on 2012-03-15 15:43

Weird... I just updated to 1068 like an hour ago.

rickeyski commented on 2012-03-15 15:38

the source 404's google apparently retracted this build as unstable_current download is now back to 1061

konart commented on 2012-03-14 20:31

Why do we need openssl098 now?

Anonymous comment on 2012-03-14 07:40

This builds are unusable at this time...Pictures are often not loaded correctly.

ridikulusrat commented on 2012-03-13 06:41

namcap report

google-chrome-dev E: ELF file ('opt/google/chrome/chrome-sandbox') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/libpdf.so') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/libffmpegsumo.so') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/nacl_helper_bootstrap') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/libppGoogleNaClPluginChrome.so') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/nacl_irt_x86_64.nexe') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/nacl_helper') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/chrome') outside of a valid path.
google-chrome-dev E: ELF file ('opt/google/chrome/PepperFlash/libpepflashplayer.so') outside of a valid path.
google-chrome-dev E: Symlink (opt/google/chrome/plugins) points to non-existing /usr/lib/mozilla/plugins
google-chrome-dev E: Symlink (opt/google/chrome/lib/libnss3.so.1d) points to non-existing /usr/lib/libnss3.so
google-chrome-dev E: Symlink (opt/google/chrome/lib/libssl3.so.1d) points to non-existing /usr/lib/libssl3.so
google-chrome-dev E: Symlink (opt/google/chrome/lib/libsmime3.so.1d) points to non-existing /usr/lib/libsmime3.so
google-chrome-dev E: Symlink (opt/google/chrome/lib/libnssutil3.so.1d) points to non-existing /usr/lib/libnssutil3.so
google-chrome-dev E: Symlink (opt/google/chrome/lib/libplc4.so.0d) points to non-existing /usr/lib/libplc4.so
google-chrome-dev E: Symlink (opt/google/chrome/lib/libnspr4.so.0d) points to non-existing /usr/lib/libnspr4.so
google-chrome-dev E: Symlink (opt/google/chrome/lib/libplds4.so.0d) points to non-existing /usr/lib/libplds4.so
google-chrome-dev E: Missing custom license directory (usr/share/licenses/google-chrome-dev)
google-chrome-dev W: File (opt/google/chrome/chrome-sandbox) is setuid or setgid.
google-chrome-dev E: Mime type handler found. Add "update-desktop-database -q" to the install file
google-chrome-dev W: Dependency desktop-file-utils detected but optional (needed for update-desktop-database)
google-chrome-dev W: Dependency libpng included but already satisfied --> I changed from libpng12 to libpng pkg in the PKGBUILD
google-chrome-dev W: Dependency included and not needed ('libxslt')
google-chrome-dev W: Dependency included and not needed ('xdg-utils')

Det commented on 2012-03-11 10:59

I just said that.

ipha commented on 2012-03-11 08:41

It appears as though it is now linked against libpng15

> ldd chrome | grep libpng
libpng15.so.15 => /usr/lib/libpng15.so.15 (0x00007fc5dd9ee000)

Det commented on 2012-03-09 13:50

Yeah. Actually it's somehow managing to use libpng15.

Anonymous comment on 2012-03-09 04:23

Why is libpng12 a dependency? I am ignoring that dependency and am having no issues.

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 2012-03-01 13:47

The 64-bit flashplugin optdep should be removed then.

Ner0 commented on 2012-03-01 09:24

Could you add openssl-compatibility as a dependency, as built-in Pepper Flash requires it?

ldd /opt/google/chrome/PepperFlash/libpepflashplayer.so
libcrypto.so.0.9.8 => /usr/lib/libcrypto.so.0.9.8 (0x00007f651962f000)

roheim commented on 2012-03-01 01:45

I deleted ~/.cache/google-chrome and ~/.config/google-chrome and it was still failing. As a workaround I went to chrome://plugins/ and disabled PepperFlash as said below.

Anonymous comment on 2012-03-01 00:29

for me it crashed upon launch. Deleting ~/.config/google-chrome/ fixed it for me.

Det commented on 2012-02-29 20:09

Dudes. Just disable the 64-bit pepper flash.

t3ddy commented on 2012-02-29 19:27

That's strange, have you tried reinstalling chrome?

Revelation60 commented on 2012-02-29 18:19

Flash (2 files) - Version: 11.2.31.109
Shockwave Flash 11.2 r31

It's enabled.

t3ddy commented on 2012-02-29 18:06

about:plugin?

Revelation60 commented on 2012-02-29 17:29

Same problem here.

t3ddy commented on 2012-02-29 17:25

Here it works fine.

Anonymous comment on 2012-02-29 16:44

Version 19.0.1055.1 not working flash (Missing Plug-In)

rickeyski commented on 2012-02-11 01:27

here is the pkgbuild with version bump and x64 md5sum http://pastebin.com/yCgsiaK0
thanks

kalpik commented on 2012-01-31 02:19

Something's gone wrong after the KDE update. When I try to download a zip, all I get is a mime-type dialog, and Chrome just hangs :/

Det commented on 2012-01-29 14:17

Noticed after creating that deb-PKGBUILD of mine that the icon symlinking part here is really messy. Something like this is much simpler: http://pastebin.com/axQCNV66

Also you'll need to use the '-f' flag with the 'lib-ln' commands or rebuilds will fail because of existing symlinks - making the specifically versioned source redundant. The 'mozilla plugin' one gets removed every time upon deleting "$pkgdir" in the beginning of the build anyway.

ruario commented on 2012-01-22 10:33

Actually this is even easier

gzip -9 "$srcdir/usr/share/man/man1/"*.1
install -Dm644 "$srcdir/usr/share/man/man1/google-chrome.1.gz" "$pkgdir/usr/share/man/man1/google-chrome.1.gz"

t3ddy commented on 2012-01-22 09:42

Thank ruario!
I thought it was packer's fault :) I was definetly on the wrong way :P

Det commented on 2012-01-21 21:27

I'll just use debs in my local PKGBUILD.

ruario commented on 2012-01-21 20:57

@Det: Seems to be bigger and marginally slower to extract because the rpm is bzip2 compressed and the deb lzma compressed. That may change in the future. It is possible to use lzma or xz compression within rpms.

ruario commented on 2012-01-21 20:55

This is the line that fails when run on 32-bit machines:

install -Dm644 "$srcdir/usr/share/man/man1/google-chrome.1" "$pkgdir/usr/share/man/man1/google-chrome.1"

because it doesn't exist. google-chrome.1.gz exists instead.

The PKGBUILD should attempt to compress any uncompressed man pages it finds and then do an install command using google-chrome.1.gz, i.e.

if [ ! -e "$srcdir/usr/share/man/man1/google-chrome.1.gz" ]; then
gzip -9 "$srcdir/usr/share/man/man1/google-chrome.1"
fi

install -Dm644 "$srcdir/usr/share/man/man1/google-chrome.1.gz" "$pkgdir/usr/share/man/man1/google-chrome.1.gz"

ruario commented on 2012-01-21 20:54

This is the line that fails when run on 32-bit machines:

install -Dm644 "$srcdir/usr/share/man/man1/google-chrome.1" "$pkgdir/usr/share/man/man1/google-chrome.1"

because it doesn't exist. google-chrome.1.gz exists instead.

The PKGBUILD should attempt to compress any uncompressed man pages it finds and then do an install command using google-chrome.1.gz, i.e.

if [ ! -e "$srcdir/usr/share/man/man1/google-chrome.1.gz" ]; then
gzip -9 "$srcdir/usr/share/man/man1/google-chrome.1.gz"
fi

install -Dm644 "$srcdir/usr/share/man/man1/google-chrome.1.gz" "$pkgdir/usr/share/man/man1/google-chrome.1.gz"

ruario commented on 2012-01-21 20:50

The reason some people have problems with the man pages and others not is down to architecture. if you run 64-bit the PKGBUILD will work, if 32-bit it won't because Google only compressed the 32-bit one.

$ bsdtar tf google-chrome-unstable-18.0.1010.0-117846.i386.rpm | grep google-chrome.1
./usr/share/man/man1/google-chrome.1.gz
$ bsdtar tf google-chrome-unstable-18.0.1010.0-117846.x86_64.rpm | grep google-chrome.1
./usr/share/man/man1/google-chrome.1

t3ddy commented on 2012-01-21 13:01

For some strange reason it doesn't extract man page, like makepkg and yaourt do.

ravicious commented on 2012-01-21 12:45

I'm using packer.

t3ddy commented on 2012-01-21 10:02

@ravicious and hagabaka
what tool are you using?

ravicious commented on 2012-01-21 09:45

Thanks hagabaka, your solution works.

Det commented on 2012-01-20 13:52

Hmh, it seems the .rpm of the latest Dev Chrome is 33.3M in size - the .deb's only 25.3M.

Extracting is also slower (~4.3s vs ~2.6s with a 6-core AMD Phenom and 7500rpm HDD).

Det commented on 2012-01-20 13:51

Hmh, it seems the .rpm of the latest Dev Chrome is 33.3M in size - the .deb's only 25.3M.

Extracting is also slower (~2.6s vs ~4.3s with 6-core AMD Phenom and 7500rpm HDD).

Det commented on 2012-01-19 22:01

@hagabaka, you need to use a *patch* for a single line? And you couldn't even use Pastebin?

t3ddy commented on 2012-01-19 21:59

Here unzipping with makepkg also extracts the man page.
What tool are you using?

I think I'm starting to understand why in chromium the man page is among source files.

t3ddy commented on 2012-01-19 21:58

Here unzipping with makepkg also extracts the man page.
What tool are you using?

hagabaka commented on 2012-01-19 21:27

I'm having the same issue as ravicious ("...google-chrome.1': No such file or directory"). The rpm file only contains google-chrome.1.gz, so unzipping the file in PKGBUILD made it work:

--- PKGBUILD.old 2012-01-19 16:26:26.000000000 -0500
+++ PKGBUILD 2012-01-19 16:22:33.000000000 -0500
@@ -40,6 +40,7 @@
ln -sv /opt/google/chrome/google-chrome "$pkgdir/usr/bin/"
ln -sv /usr/lib/mozilla/plugins "$pkgdir/opt/google/chrome"
mv -v "$pkgdir/opt/google/chrome/google-chrome.desktop" "$pkgdir/usr/share/applications"
+ gunzip "$srcdir/usr/share/man/man1/google-chrome.1.gz"
install -Dm644 "$srcdir/usr/share/man/man1/google-chrome.1" "$pkgdir/usr/share/man/man1/google-chrome.1"

mkdir "$pkgdir/opt/google/chrome/lib"

Anonymous comment on 2012-01-19 20:33

@ravicious worked fine for me using packer...something must have gone wrong before this...maybe try again with packer or download the pkgbuild and use makepkg

Det commented on 2012-01-19 20:15

Works fine with makepkg/yaourt here.

ravicious commented on 2012-01-19 20:12

==> Actual installation
`/tmp/packerbuild-1000/google-chrome-dev/google-chrome-dev/pkg/usr/bin/google-chrome' -> `/opt/google/chrome/google-chrome'
`/tmp/packerbuild-1000/google-chrome-dev/google-chrome-dev/pkg/opt/google/chrome/plugins' -> `/usr/lib/mozilla/plugins'
`/tmp/packerbuild-1000/google-chrome-dev/google-chrome-dev/pkg/opt/google/chrome/google-chrome.desktop' -> `/tmp/packerbuild-1000/google-chrome-dev/google-chrome-dev/pkg/usr/share/applications/google-chrome.desktop'
install: cannot stat `/tmp/packerbuild-1000/google-chrome-dev/google-chrome-dev/src/usr/share/man/man1/google-chrome.1': No such file or directory
==> ERROR: A failure occurred in package().
Aborting...
The build failed.

Anonymous comment on 2012-01-03 16:49

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.

Det commented on 2011-12-16 14:04

A crappy version. Hangs all the time.

Det commented on 2011-12-13 07:46

It's not yet in the blog but .6 has already been released.

brycec commented on 2011-12-10 08:33

MD5 for amd64 17.0.963.2-r113542
0497205a42b53cf6e4def803306f58d1 google-chrome-unstable_current_amd64.deb

brycec commented on 2011-12-10 08:29

i386 deb seems to have been bumped to .2
dpkg -I google-chrome-unstable_current_i386.deb| grep Version
Version: 17.0.963.2-r113542
024678c5263356d0ae6a7e6a1ca8a147 google-chrome-unstable_current_i386.deb

Anonymous comment on 2011-12-09 09:38

yep no extensions will install

t3ddy commented on 2011-12-09 09:24

I've got that problem too.
It could be related to this issue: http://code.google.com/p/chromium/issues/detail?id=106519

Whitie commented on 2011-12-09 08:25

The last version does not install any extension. Has anyone else this problem?

Anonymous comment on 2011-11-08 00:06

new x86_64 md5: b961c82a6dbf1887d5c3fcb873bbeeeb

Anonymous comment on 2011-11-04 08:19

new x86 md5: ae0381e5fcd4285f203f5450228b9d06

michael.manley commented on 2011-11-02 08:15

x86_64 md5sum changed to 75a837ce451a6b64dc3c8c5ed629af2a Version is now 16.0.912.21.

Det commented on 2011-10-28 15:42

@vladimir, it means that the package needs to be updated.".."

Anonymous comment on 2011-10-28 06:43

x86 needs the md5sum changed to 0fae792cac99a7c1bb10f7c89959fa1e to actually install. Otherwise works fine

michael.manley commented on 2011-10-26 02:44

x86_64 needs the md5sum changed to d37efc28ccc3fba673fb6d5e16354246 to actually install. Otherwise works fine

Anonymous comment on 2011-10-03 19:06

Seems to work for me: Click once to star. Click again to bring up the popup. You can then remove the star through the popup.

Anonymous comment on 2011-10-03 15:09

I've noticed that all of the google-chrome-* packages in AUR appear to have the following problem:

Clicking the Star icon in the omnibox does not show a the Bookmark popup panel (that lets you choose the folder to store the bookmark in). You also can't click the star to remove the bookmark once it has been added. Using the .deb directly in another Linux doesn't seem to have this problem, however.

Anonymous comment on 2011-10-02 13:30

# man PKGBUILD

It is also possible to change the name of the downloaded file, which is helpful with weird URLs and for handling multiple
source files with the same name. The syntax is: source=('filename::url').

How about

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

and

ar -xv google-chrome-${_channel}_current_${pkgver}_${_arch}.deb

Similarly for other google-chrome packages as well.

Det commented on 2011-09-29 08:04

Then it's a pacman bug and you should report it to the actual pacman developers (while the thing is still a release candidate).

Anonymous comment on 2011-09-28 15:53

@Det:

keshav_pr@my-system ~/Desktop/google-chrome-dev % makepkg
==> ERROR: pkgver is not allowed to contain colons, hyphens or whitespace.

makepkg from pacman 4.0.0 RC2.

Det commented on 2011-09-24 08:56

@skodabenz, or you should just start using a better AUR tool (or at least notify the maintainer).

("pkgver='15.0.874.21' # Check [..]" should be enough anyway. It doesn't need to be separated into 2 different lines.)

Anonymous comment on 2011-09-24 07:24

ERROR: pkgver is not allowed to contain colons, hyphens or whitespace.

Should change

pkgver=15.0.874.21 # Check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Dev%20updates

to

## Check for new Linux releases in: http://googlechromereleases.blogspot.com/search/label/Dev%20updates
pkgver="15.0.874.21"

Anonymous comment on 2011-09-11 08:27

15.0.874.5 amd64 md5sum: cc46f5bb134fc50a4d6cd9c43a44b175

Anonymous comment on 2011-08-25 11:00

Bug is confirmed as currently fixed.

Anonymous comment on 2011-08-20 07:19

Bug is reportedly fixed in current build. Should be good to go on the next update.

sl1pkn07 commented on 2011-08-19 02:48

important "bug"

dont close tabs (only with cntl+W)

see
http://code.google.com/p/chromium/issues/detail?id=93086

eyenx commented on 2011-08-17 19:06

google-chrome-dev 15.0.854.0-1 (i386) md5sum: 133b8a76d58e1b3786459bf06dbf1a10

eyenx commented on 2011-08-17 08:28

15.0.849.0-1 i386 md5sum 133b8a76d58e1b3786459bf06dbf1a10

Anonymous comment on 2011-08-16 23:29

15.0.849.0-1 (x86_64) md5sum 3a18d3c4d5753b1b225f03851e847071

samuvuo commented on 2011-08-09 03:33

14.0.835.18: i386 m5sum 5c3c78a0736c3069e73f13056e93e8e6

magicfish1990 commented on 2011-08-09 00:51

14.0.835.29-r95190 x86_64 md5sum efaaa80043d6cd8a639997dea7178866

Anonymous comment on 2011-08-04 18:40

Nevermind. Latest nvidia update fixed it.

Anonymous comment on 2011-08-01 18:07

Hmm.. is anyone else having trouble viewing HTML5 content? All of a sudden, HTML5 pages are just empty and blue now.

Det commented on 2011-07-30 08:45

Seems like it. His last update was 21th (google-chrome-beta). Let's give him time.

In the meantime here's an updated PKGBUILD for those making use of it: http://pastebin.com/S6Q7NJcX

ravicious commented on 2011-07-29 22:50

Is t3ddy on a holiday or something?

ravicious commented on 2011-07-29 20:37

Is t3ddy on holiday or something?

Anonymous comment on 2011-07-28 22:27

4f11bc2a10b8dab4709c022ba462ae63 64-bit checksum for 14.0.835.0

Anonymous comment on 2011-07-14 19:12

Ah, found the problem. Didn't realize there were multiple "exec" lines in the .desktop file. Whoops. That'll teach me to jump to the bottom of files.

Det commented on 2011-07-13 18:48

I can't reproduce that here. Changing the .desktop file in /usr/share/applications was enough. Maybe you didn't save the file or re-installed Chrome afterwards (replacing the .desktop file)? Or perhaps you just mistyped the flag.

Anonymous comment on 2011-07-13 17:54

Got my hopes up. Unfortunately it's not in there :(

Det commented on 2011-07-12 23:53

It may also be in .local/share/applications/.

Anonymous comment on 2011-07-12 15:35

Sorry, I forgot to mention I'm running Gnome 3. I've been starting Chrome through the built-in application launcher. I assumed it was pulling from /usr/share/applications, but changing the google-chrome.desktop exec line did nothing. I guess I could run it manually...

Det commented on 2011-07-12 00:49

Just either add that after the menu item's exec line or start Chrome manually with that in the end.

Anonymous comment on 2011-07-11 19:37

Does anyone know how to change the startup flags/arguments? I'm trying to get the browser to start with --purge-memory-button

Det commented on 2011-06-17 13:39

Nobody's hurt or anything :).

t3ddy commented on 2011-06-17 13:02

you're right, I'm a bit off these days :S

Det commented on 2011-06-17 12:35

The .desktop file says "Icon=google-chrome" - so no icons are needed in the root of /usr/share/icons.

Anonymous comment on 2011-06-17 00:51

Dev has been updated to version 14, this needs to be updated.

t3ddy commented on 2011-06-15 12:41

I think this line:
cp -v "$pkgdir/opt/google/chrome/product_logo_256.png" "$pkgdir/usr/share/icons/google-chrome.png"
is needed by .desktop

for the other two, I've replaced them

Det commented on 2011-06-14 21:23

Hey, again. These lines:

install -d "$pkgdir/usr/share/icons/hicolor/${i}x${i}/apps"
cp -v "$pkgdir/opt/google/chrome/product_logo_${i}.png" "$pkgdir/usr/share/icons/hicolor/${i}x${i}/apps/google-chrome.png"

could be replaced with just:

install -D (-v) "$pkgdir/opt/google/chrome/product_logo_${i}.png" "$pkgdir/usr/share/icons/hicolor/${i}x${i}/apps/google-chrome.png"

Also is this really needed?:

cp -v "$pkgdir/opt/google/chrome/product_logo_256.png" "$pkgdir/usr/share/icons/google-chrome.png"

t3ddy commented on 2011-06-14 21:13

I've fixed the formatting, is it working for you now?

anish commented on 2011-06-14 20:34

==> Validating source files with md5sums...
google-chrome-unstable_current_amd64.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!

This happens to me with every single update, every single time

t3ddy commented on 2011-06-11 15:54

I don't think it's something crucial, although I could fix it... I'll see

anish commented on 2011-06-11 01:04

I think you should fix the formatting for the sources/md5sum array, should help aur parse it better instead of showing
&&
http://dl.google.com/linux/direct/google-chrome-unstable_current_
md5sums=d7a3db4e0ffba36c04fa06e42ae7d3fc.deb

Kernald commented on 2011-05-18 09:41

New md5sum for the amd64 : 2f882a70a3f97cbc5654af6de0e73965

Det commented on 2011-05-13 18:39

It's not yet in the blog but 13.0.761.0 has been released for all platforms (or, well, at least it is for Windows and the md5sums are different with the debs).

Anonymous comment on 2011-05-05 04:33

I have removed libjpeg6 for few days and Chrome worked fine during this time.
libjpeg-turbo is required by other packages anyway.

Det commented on 2011-05-04 19:46

I think so.

t3ddy commented on 2011-05-04 19:05

So I just have to remove libjpeg6 and nothing else, right?

t3ddy commented on 2011-05-04 19:04

So the change to do is from libjpeg6 to libpng?

Det commented on 2011-05-04 18:28

I removed libjpeg6 and all works fine. Libjpeg-turbo doesn't seem to be required either:

└┌(%:~)┌- lsof /usr/lib/libjpeg.so
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
tint2 920 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
conky 929 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
yakuake 938 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
kded4 958 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
kglobalac 962 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
knotify4 968 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2

E: Lol, no, forget this part I said about libpng12. Chrome doesn't start without it (so apparently it uses both).

Det commented on 2011-05-04 18:26

I removed libjpeg6 and all works fine. Libjpeg-turbo doesn't seem to be required either:

└┌(%:~)┌- lsof /usr/lib/libjpeg.so
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
tint2 920 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
conky 929 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
yakuake 938 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
kded4 958 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
kglobalac 962 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2
knotify4 968 det mem REG 8,1 246976 787367 /usr/lib/libjpeg.so.8.0.2

Also the libpng dependency should be v14 ('libpng') instead of the old v12 ('libpng12'):

└┌(%:~)┌- lsof /usr/lib/libpng.so|grep chrome
chrome 1240 det mem REG 8,1 168632 789782 /usr/lib/libpng14.so.14.5.0
chrome 1244 det mem REG 8,1 168632 789782 /usr/lib/libpng14.so.14.5.0
chrome 1340 det mem REG 8,1 168632 789782 /usr/lib/libpng14.so.14.5.0

t3ddy commented on 2011-05-04 16:46

:D
I think of switching to libjpeg-turbo with the next release

Anonymous comment on 2011-05-04 16:15

Yep. (Looks at libjpeg-turbo's Required By:) Doesn't everyone?

Revelation60 commented on 2011-05-04 12:23

I get a "missing plug-in" error when trying to view a flash movie. This only happens in the latest version of this package, before that it was fine.

t3ddy commented on 2011-05-03 19:39

Do you have libjpeg-turbo installed?

Anonymous comment on 2011-05-03 14:34

For me removing libjpeg6 works too.

t3ddy commented on 2011-05-02 16:26

In the .deb file, libjpeg6 is listed as dependency, although I've tried removing it and rebuilding the chrome package, everything seems fine.
If someone else can make some tests, they'll be appreciated.

Anonymous comment on 2011-05-02 03:40

Does Chrome still depend on libjpeg6?
I thought Chrome switched to libjpeg-turbo since 11? http://code.google.com/p/chromium/issues/detail?id=48789

magicfish1990 commented on 2011-04-29 09:27

12.0.742.12 dev
amd64 md5sum = d00818bf674f4f9122d020a26940930e

sjakub commented on 2011-04-29 08:01

If someone can't figure out things like current version number and MD5 sum, maybe they shouldn't be using the development version?

trizen commented on 2011-04-29 07:55

pkgver=12.0.742.12
[...]
_arch='i386' && md5sums=('328bc48c8c18290c631595091af2ded5')

Anonymous comment on 2011-04-29 01:32

google-chrome-unstable_current_i386.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!
==> ERROR: Makepkg was unable to build google-chrome-dev.

letroll commented on 2011-04-25 19:10

md5sum pour 64bits 33c0e4593a81909e4487477b7d42c744

Anonymous comment on 2011-04-24 19:00

the md5 check for i686 version is '60d51842347f25709d8a8d321587d1aa'

Anonymous comment on 2011-04-21 20:43

changed to md5 check to '2A497BF5B8F04C3DF1CC23D336FBAC6F' for the amd64 version and the build to 12.0.742.0 and everything works

Det commented on 2011-04-13 11:33

Btw. the .install file's "post_remove()" line could just be "post_install" instead of actually repeating "post_install" (as with the others Chromes too).

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

t3ddy commented on 2011-03-25 20:55

I've added a bug report

Det commented on 2011-03-25 13:59

Good point, sjakub".."

sjakub commented on 2011-03-25 13:58

I think the right way is opening a bug request instead of emailing people...

t3ddy commented on 2011-03-25 13:54

I don't have testing enabled, anyway when pciutils-3.1.7-4 will hit core, I'll change the pkgbuild :)
I think I'll send an e-mail to the maintainer, if none has already done it.

Det commented on 2011-03-25 13:35

pciutils-3.1.7-4 provides libpci.so. The pciutils maintainer (Tobias Powalowski) should probably be emailed to add "provides=(libpci)" to the PKGBUILD or something.

Anonymous comment on 2011-03-25 04:43

md5 check failed

jnbek commented on 2011-03-04 03:08

Grr, the md5s for x64 is broken...

trizen commented on 2011-03-02 06:06

pkgver=11.0.686.1

i386:
md5sums=('f6a07a51748a17b13b76f615c25b7ae1')

Anonymous comment on 2011-03-02 00:08

pkgver=11.0.686.0

x86_64:
md5sums=('17cae321f43a34bf0a94fb8c0be88ede')

trizen commented on 2011-03-01 07:05

pkgver=11.0.686.0
......
i386:
md5sums=('442ef1c5bc29c0d91bc2faa1671433ad')

Det commented on 2011-02-19 12:49

@LookTJ, they kinda already are...

LookTJ commented on 2011-02-19 09:22

separate the architectures' md5

x86_64:
md5sums=('739366fe80400cc5d574ed3966a8a8ff')

emhs commented on 2011-02-19 00:10

New update fixed it. Thanks, all.

Anonymous comment on 2011-02-18 23:58

@emhs, no, the md5 for i686 is correct - 3ec9ea320a357e0357968aad2ea604f6 google-chrome-unstable_current_i386.deb

emhs commented on 2011-02-18 22:56

tanguyr, I'm running i686. Is the md5 wrong there too?

Det commented on 2011-02-18 13:15

My pleasure.

trizen commented on 2011-02-18 10:23

Thanks mister teacher.

Det commented on 2011-02-18 09:58

It's "latest" in English.

trizen commented on 2011-02-18 09:53

Last version is: 11.0.672.2

Anonymous comment on 2011-02-18 08:09

md5 for x86_64 is incorrect, it should be 739366fe80400cc5d574ed3966a8a8ff

emhs commented on 2011-02-18 01:32

I keep getting checksum failures when trying to upgrade from 10.634.0 to 10.648.82. Any guesses what that's about?

sjakub commented on 2011-02-17 19:31

I have flashplugin-prerelease 10.3.162.29-4 and no problems or warnings. Check your version of the package.

Anonymous comment on 2011-02-17 19:29

Not sure of the exact version off the top of my head, but it's whatever the standard version in the repos is.

Det commented on 2011-02-17 19:21

Nope. What flash are you using then?

Anonymous comment on 2011-02-17 16:06

Right. So is anybody else seeing this message every time a you load a page with flash?

Det commented on 2011-02-17 16:04

@flammenwurfer, but that was not what was asked.

Anonymous comment on 2011-02-17 15:39

Ok, well I'm running it on 64 bit arch. That better?

sjakub commented on 2011-02-17 15:24

No it's not "64 bit version". It's both, depending on the architecture you make this package on.

Anonymous comment on 2011-02-17 15:18

It is the 64bit version.

Raymondcal commented on 2011-02-17 15:15

Hi !
Is this the 32 or 64bits version ?

Anonymous comment on 2011-02-03 05:18

Is anybody else getting a notification that the flash plug-in is blocked because it is out of date?

Det commented on 2011-01-21 21:28

@scio, that's why you weren't the one doing it :).

scio commented on 2011-01-21 19:51

@Det: sorry my mistake on the order, but the idea still stands.
If another application is looking for .49 it will find (the now symlinked) .50. It's just not a good idea no matter which way you are doing it.

Det commented on 2011-01-21 18:52

@scio, I don't understand where did you get the idea that the symlink would point to the old version. It points from the _old_ one (.so.49) to the _new_ one (.so.50) and not the other way around.

When and if the new Chrome is built for the new libavutil (.so.50) it will just read that one and never even know about the symlink.

Weegee commented on 2011-01-21 17:56

Problem solved! I just had to delete ~/.local/share/applications/google-chrome.desktop ... ^^'

scio commented on 2011-01-21 15:12

@Det: My advice the is opposite of being short-sighted, it is preventing future problems. Yes, it can be used short term to fix an upgrade problem like yours, but doing it for a package like this that is updated all the time seems like a bad idea. If he makes the symlink now and uses the browser for a day or two without problem he will most likely forget he made the symlink. The next version comes which actually uses the correct version, but you symlink points to the old one and segfault.

This is a dev version of a browser, you should not be making symlinks of system libraries to make it work.

Det commented on 2011-01-21 14:44

And that mattoufoutu's issue seems to be the cause of updated ffmpeg (which increased the libraries' versions) because Chrome uses the system ffmpeg. There's really nothing to do expect to wait for the next Dev Chrome and see if it's built for the new libavutil.

@scio, don't be so short-sighted. There are cases when symlinking is the only available solution and this is one of them (dunno how much libavutil even affects Chrome, though). Another case that happened to me was during the openssl/libgcrypt transition (back in April 2010). Pacman was rendered totally useless because it was built for the former libraries and I (of course) had removed the former versions of libgcrypt/openssl from the pacman cache with 'pacman -Sc' (so I neither could I manually place the files from the tarballs). Well, my browser wasn't working either so fetching the former versions from the internets was also out. So what did I figure to do next? Ah yes, _symlinking_, that's the one. So I temporarily symlinked the new libs to the old ones, updated pacman (along with everything else) and finally removed the symlinks.

Not something you'd do in normal circumstances but instead something that you might occasionally do when all other options have gone out.

scio commented on 2011-01-21 14:04

@mattoufoutu: that is a terrible way to fix that problem. That means that the package needs to be rebuilt against your new library. Symlinking a library to a different version number will only cause problems.

Det commented on 2011-01-21 14:02

Maybe some setting or cache in chromium's or xdg-util's side? Eg. you could try moving out the Chromium settings folder and then checking whether the problem persists ~/.config/chromium?

The installation files should all be automatically replaced upon updating (leaving no traces of the former installation folder path) so reinstalling this thing shouldn't help for that matter.

Anonymous comment on 2011-01-21 14:00

At startup, google-chrome is looking for the libavutil library in "/usr/lib/libavutil.so.49",
but the correct file is at "/usr/lib/libavutil.so.50", creating a symlink solves the problem.

t3ddy commented on 2011-01-21 12:57

It seems all right then.
I don't know what could cause the problem.

Weegee commented on 2011-01-20 21:51

lrwxrwxrwx 1 root root 32 20. Jan 22:48 /usr/bin/google-chrome -> /opt/google/chrome/google-chrome

t3ddy commented on 2011-01-20 21:27

.desktop is fine
"ls -l /usr/bin/google-chrome" what gives you?

Weegee commented on 2011-01-20 21:19

Exec=/opt/google/chrome/google-chrome %U

t3ddy commented on 2011-01-20 21:05

And what about the Exec line in the .desktop?

Weegee commented on 2011-01-20 20:35

Nah, I've built the package with makepkg (using the tarball), but it still didn't work :/

t3ddy commented on 2011-01-20 19:50

@Weegee
maybe you've got the same problem that had sjakub i.e. yaourt cache
I've decided to revert back to the original /opt/google/chrome instead of /opt/google-chrome because it seems that google's apps are all installed into /opt/google

Weegee commented on 2011-01-20 19:47

By the way, why doesn't "Make Google Chrome My Default Browser" actually make Google Chrome my default browser? If I start Chrome from within the terminal, I just get "which: no google-chrome in (/opt/google-chrome)" when I click on the button in the settings tab :(

t3ddy commented on 2011-01-20 10:43

Updated to 10.0.642.2.

I've decided to revert back to the original directory, so now it's again in /opt/google/chrome.
I don't think that anything cares where chrome is installed, but if also other google's apps are into /opt/google, probably also chrome should be there.

I've also removed the directory /usr/lib from the pkgbuild since nothing was put in it and the optdependency moonlight has been removed, since I think is more than a year that a package with that name doesn't exists.

trizen commented on 2011-01-20 03:27

10.0.642.2 is out

t3ddy commented on 2011-01-19 21:22

I'll do what you prefer, since for me it makes no difference at all.

Det commented on 2011-01-19 16:48

@sjakub, maybe you shouldn't post everything so quickly so that you wouldn't need to delete these posts afterwards either. In addition you would've seen t3ddy's response to you where he already told you why he changed the install dir.

As for my stance about this, atm, you can't find a single "thing" that assumes Google Chrome is installed to "/opt/google". Nothing looks for it in there and nothing cares it's in there. I originally made the change because I had no idea that all Google stuff would/should be installed there. I just figured it was redundant to have a folder for Chrome where inside you got another folder that then would have the actual program - as such I changed the dir.

Frankly, I don't think anybody even cares where Chrome is installed (as long as it just works) but it wouldn't be all bad idea to keep all the Google stuff in a single directory - as upstream wants.

T3ddy's decision.

sjakub commented on 2011-01-19 16:19

I am wondering if it is going to cause problems. All other Google packages go to /opt/google/*. Also, since this is standard google's location, some things might assume it's there.
Why did you want it Det?

t3ddy commented on 2011-01-19 16:14

Det has made the proposal, personally I had nothing counter and none else has said anything, so the change has been made.

sjakub commented on 2011-01-19 15:46

Sigh. No, you're right. I reinstalled everything and it was still wrong, but it was due to yaourt's caching :/ When I grabbed the tarball of this AUR package and did everything manually, just in case, it was fine.
Sorry for that :(

As a side note - why do you move it from /opt/google/chrome to /opt/google-chrome?

t3ddy commented on 2011-01-19 15:32

@sjakub
Are you absolutely sure that "grep Exec /usr/share/applications/google-chrome.desktop" returns something different from "Exec=/opt/google-chrome/google-chrome %U" like it does on my system?

sjakub commented on 2011-01-19 15:15

@Det, the PKGBUILD is BROKEN right now. It provides a desktop file that points to a location that doesn't exist. In your opinion this is correct and doesn't require fixing?

Weegee commented on 2011-01-18 19:12

@ Det

of course I changed "/google/chrome/" to "/google-chrome/", however it was still broken (and I was confused, hehe), so now I'm going to use your script (hope it'll work :<)

Det commented on 2011-01-18 10:05

@sjakub, a single post is quite enough and the PKGBUILD is not "BROKEN" nor does it need "fixing" (atm). Also there has already been talk about the .desktop file :).

@Weegee, of course it does. You just need to change the "/google/chrome/" indications to "/google-chrome/" (since that's where this package is currently installed to) and realize that "the google-chrome script" part needs to be redone every time the main Chrome script is overwritten (Chrome is updated) :).

Here's the step-by-step instructions again (tweaked a bit): http://aur.pastebin.com/TYJW0LGZ

I also made a script for doing this (mostly out of boredom - took like an hour so I guess it was worth it). You (of course) need to "chmod +x" it and run it as root (but you probably wanna _check_ the script at first before doing so - nor would I blame you if you did). This also works with iron and chromium-browser (you can manually choose which one you like (only one at the time, though)): http://aur.pastebin.com/8B7R9m48

Det commented on 2011-01-17 22:07

@sjakub, a single post is quite enough and the PKGBUILD is not "BROKEN" nor does it need "fixing" (atm). Also there has already been talk about the .desktop file :).

@Weegee, of course it does. You just need to change the "/google/chrome/" indications to "/google-chrome/" (since that's where this package is currently installed to) and realize that "the google-chrome script" part needs to be redone every time the main Chrome script is overwritten (Chrome is updated) :).

Here's the step-by-step instructions again (tweaked a bit): http://aur.pastebin.com/TYJW0LGZ

I also made a script for doing this (mostly out of boredom - took like an hour so I guess it was worth it). You (of course) need to "chmod +x" it and run it as root (but you probably wanna _check_ the script at first before doing so - nor would I blame you if you did). This also works with iron and chromium-browser (you can manually choose which one you like (only one at the time, though)): http://aur.pastebin.com/TusAWDGw

t3ddy commented on 2011-01-17 09:11

this line fixes what you say: sed -i "s|google/chrome|google-chrome|g" $(grep -Rls "/google/chrome" *)
and is already in the pkgbuild

det has suggested the directory change, none has said anything in contrast, so it has been done

t3ddy commented on 2011-01-17 09:08

this line fixes what you say: sed -i "s|google/chrome|google-chrome|g" $(grep -Rls "/google/chrome" *)
and is already in the pkgbuild

sjakub commented on 2011-01-16 23:26

Here is fixed PKGBUILD, with working .desktop file, that installs chrome into /opt/google/chrome: http://aur.pastebin.com/03QJhsfd

sjakub commented on 2011-01-16 23:23

fresh24: Try fixing the .desktop file, maybe Gnome doesn't show the entry because of incorrect Exec line?

sjakub commented on 2011-01-16 23:23

fresh24: Try fixing the .desktop file, maybe Gnome doesn't who the entry because of incorrect Exec line?

sjakub commented on 2011-01-16 23:18

This package is BROKEN right now.
It provides .desktop file that tries to exec /opt/google/chrome/google-chrome, which does not exist.
You are actually breaking it by moving opt/google/chrome to opt/google-chrome in PKGBUILD (what is the reason for doing that?).
If you move things around, please modify the .desktop file accordingly. I don't see a reason for changing the location from /opt/google/chrome,
since all google programs go into /opt/google (picasa, talkplugin). So instead of moving things around you could just leave them
in the default path, and the provided .desktop file would work properly.

Weegee commented on 2011-01-16 19:16

So ... how do I fix the libjpeg issue? Neither the google-chrome-fix package nor the workaround posted by master do work anymore ...

t3ddy commented on 2011-01-16 10:57

Already tried, but ever the same result.

Anonymous comment on 2011-01-15 18:36

Working fine here. Try cleaning out the build directory?

t3ddy commented on 2011-01-15 18:21

I've noticed that too, but I can't figure out the reason.
Changing the category seems to have no effect, so I don't know...

fresh24 commented on 2011-01-15 11:45

Google Chrome doesn't appears in the gnome-menu anymore, doesn't work with gnome-do and as docky launcher.
It seems, that there is a issue with the .desktop file.

Det commented on 2011-01-12 07:00

@Wicked0ne Yeah, the Dev Channel Chrome was updated to 10.0.634.0, which also changed the md5sums (the links never change between any of the three channels ('stable', 'beta' and 'dev') when updating). The new ones are '19aed3f0f670666c6bf92eb1804883f6' for amd64 Chrome and for i386 Chrome: '9df80987b6302b8a805401feb87da0c7'.

Anonymous comment on 2011-01-12 02:58

bad md5sums here.


==> Validating source files with md5sums...
google-chrome-unstable_current_amd64.deb ... FAILED
==> ERROR: One or more files did not pass the validity check!
error: Build failed

t3ddy commented on 2011-01-10 07:48

That link is provided by the pkgbuild to ease the launch of chrome from command line, but by default the .desktop points to /opt/google/chrome/google-chrome, so I can't understand why I should change it to /usr/bin/google-chrome (that actually is a simlink), instead of /opt/google-chrome/google-chrome (the file linked).

sjakub commented on 2011-01-10 02:58

Wherever chrome is installed, it provides /usr/bin/google-chrome and that binary should be specified in .desktop file...

$ pacman -Qo /usr/bin/google-chrome
/usr/bin/google-chrome is owned by google-chrome-dev 10.0.628.0-2

Det commented on 2011-01-09 19:07

Right, of course. Gotcha :). That my way worked only because the .desktop file _name_ is "google-chrome.desktop".

t3ddy commented on 2011-01-09 17:16

No, since with grep I'm searching files that contains /opt/google/chrome, to change it into /opt/google-chrome

Det commented on 2011-01-09 17:04

Oh yeah, that's better, but shouldn't it be: sed -i "s/google\/chrome/google-chrome/g" $(grep -Rl "/google-chrome" *)

t3ddy commented on 2011-01-09 16:35

the .desktop gets changed by: sed -i "s/google\/chrome/google-chrome/g" $(grep -Rl "/google/chrome" *)
since I prefer to change all the files that has this problem

Det commented on 2011-01-09 16:18

@t3ddy, actually, as I suggested and you applied to your PKGBUILD, it's just "/opt/google-chrome" now - it's simpler that way :).

The .desktop file can (and should be) changed with e.g.: sed -i 's|google/chrome|google-chrome|' "$pkgdir/opt/google-chrome/google-chrome.desktop"

t3ddy commented on 2011-01-09 13:21

The .desktop points to /opt/google-chrome/google-chrome since is there that chrome is installed

sjakub commented on 2011-01-09 12:47

The .desktop file contains path to /opt/google/chrome/google-chrome, but chrome is now installed in /usr/

t3ddy commented on 2011-01-08 20:32

Ok, I've fixed the $pkgdir :)

Det commented on 2011-01-08 17:46

Thanks for the update but forget something did you around the "$pkgdir" variable before your little sed hack? Something like.. quotes (") :)?

In addition the "then" part there might as well be on the same line like this: http://aur.pastebin.com/qNkFQ1Jj

scio commented on 2011-01-08 14:24

This line doesn't work for me:
_arch='i386' && md5sums=('a677c4e4d9a2ef889a9f5a254b809408')

Not quite sure why, looks like valid bash but I had to split it to 2 lines to get the deb to download correctly.

_arch='i386'
md5sums=('a677c4e4d9a2ef889a9f5a254b809408')

t3ddy commented on 2011-01-08 11:50

Updated, sorry but I was away :)

Det commented on 2010-12-29 16:31

...and "optdepends=(${optdepends[*]} 'flashplugin')" could just be "optdepends+=('flashplugin')" =).

Det commented on 2010-12-24 16:50

Oh, and the "check for new releases in" comments of mine next to the pkgvers there could actually point straight to the labeled stable/beta/dev channel updates - to only show the updates of that specific label. E.g. the stable channel: http://googlechromereleases.blogspot.com/search/label/Stable%20updates

trizen commented on 2010-12-23 22:27

@xonecas - libjpeg6 is in AUR so you need a AUR Package Manager to resolve all dependencies automatically, as yaourt, packer, clyde, trizen, etc...

Det commented on 2010-12-23 11:35

Augh. Seems I put the same Pastebin link for both 'google-chrome' and 'google-chrome-beta'. I'll just repost this whole thing to not bury it underneath that flash discussion (now also includes the 'flashplugin' changes):

I've always wanted to say this in the AUR comment sections :) : Please pull my stuff:

google-chrome: http://aur.pastebin.com/FkxUZnYu

google-chrome-beta: http://aur.pastebin.com/dpAjUZRN

google-chrome-dev: http://aur.pastebin.com/bSygGHt8

(It's just simplifications to the PKGBUILDs.)

t3ddy commented on 2010-12-23 11:05

I'll consider merging that changes when the next release came out. :)

Det commented on 2010-12-22 16:43

I don't think it's released yet. Still beta.

scio commented on 2010-12-22 16:41

@Det: That would probably be it, I thought after the release of the adobe 64-bit player, chrome was updated as well. It appears that is not the case.

Det commented on 2010-12-22 16:36

Might this be it :)?: "There is no bundled Adobe Flash Player plug-in for 64-bit Linux." (http://googlechromereleases.blogspot.com/2010/03/dev-channel-update_30.html)

Please don't tell me you _have_ a 64-bit system (since I do). If you don't a solution could be something like:

[ "$CARCH" = "x86_64" ] && optdepends=('flashplugin' 'moonlight' 'gecko-mediaplayer')

scio commented on 2010-12-22 16:27

@Det: yes flash works fine for me.

Det commented on 2010-12-22 16:24

Yeah, as I said, I tried the (latest) Stable/Beta/Dev Channel Chromes and it didn't work with any of them. Does it work for you or anybody else? I mean my system is (of course) up-to-date and every time I tried with a _clean_ settings directory (.config/chromium).

scio commented on 2010-12-22 16:02

@Det: I'm not sure what your problem is, as of 5.0.375.9[1] flash is built in and enabled by default. There is no need to add command line arguments (which have also been replaced with about:flags).

You don't need to do anything to make flash work. Install google-chrome-dev, and it should work out of the box. If your version of chrome is using flashplugin, you either have a very old version or something is not built correctly.

[1]http://googlechromereleases.blogspot.com/2010/04/dev-channel-update_12.html

Det commented on 2010-12-22 15:34

@scio, I didn't really get the ending of your message and the other part was mostly repeating your former one but ok, then. I get that. It's just that now that I try viewing (for the second time) anything 'flashy', I just keep getting the "Missing Plug-in" message (whether Chrome was started with "--enable-internal-flash" or not). I get this with the Beta and Stable Channel Chrome's too.

So if you could just elaborate on how Chrome's builtin Flash works instead of saying how it does "NOT" work, then that'd be great.

scio commented on 2010-12-22 15:04

@Det: google-chrome-dev has built in flash, it does NOT use flashplugin and thus doesn't need flashplugin as an optdepends.

I can't say for the others, I don't remember which versions have it or not. You are posting this on the google-chrome-dev package so that is what I am referring to.

Det commented on 2010-12-22 14:46

I've always wanted to say this in the AUR comment sections :) : Please pull my stuff:

google-chrome: http://aur.pastebin.com/sJpsjpnH

google-chrome-beta: http://aur.pastebin.com/sJpsjpnH

google-chrome-dev: http://aur.pastebin.com/VR9nR4yV

(It's just simplifications to the PKGBUILDs.)

E: After all, flash does not come with the Linux version of Chrome, so flashplugin can be listed as an optional dependency. The installation directory can also be set as just "/opt/google-chrome" instead of "/opt/google/chrome".

E2: @scio, so it _does_ indeed? At least youtube didn't work when I tried it.

scio commented on 2010-12-22 14:37

@Det: flash does come with the linux version of chrome, it does NOT use flashplugin.

Det commented on 2010-12-22 14:26

I've always wanted to say this in the AUR comment sections :) : Please pull my stuff:

google-chrome: http://aur.pastebin.com/WBxE4kSz

google-chrome-beta: http://aur.pastebin.com/zyUTz8cP

google-chrome-dev: http://aur.pastebin.com/pbNwHgMv

(It's just simplifications to the PKGBUILDs.)

E: After all, flash does not come with the Linux version of Chrome, so flashplugin can be listed as an optional dependency. The installation directory can also be set as just "/opt/google-chrome" instead of "/opt/google/chrome".

Det commented on 2010-12-22 13:53

I've always wanted to say this in the AUR comment sections :) : Please pull my stuff:

google-chrome: http://aur.pastebin.com/KSufwDE4

google-chrome-beta: http://aur.pastebin.com/xvh8UWpS

google-chrome-dev: http://aur.pastebin.com/VqyXtWQa

(It's just simplifications to the PKGBUILDs.)

scio commented on 2010-12-10 13:13

@x-demon: You could try asking bdheeman, he maintains the chromium packages.

Anonymous comment on 2010-12-10 12:49

anyone want to maintain chrome packages? Please write to me at lex.io

LookTJ commented on 2010-12-03 23:32

ar: google-chrome-unstable_current_i386.deb: Malformed archive
Aborting...
error: Build failed

fault of Google?

Anonymous comment on 2010-11-20 15:37

/usr/bin/google-chrome: error while loading shared libraries: libXtst.so.6: cannot open shared object file: No such file or directory

libxtst as dependency solves it

Det commented on 2010-11-17 17:26

Now that I remember, the first is true. About the second one, I never said you didn't understand what the optdepends are for :).

scio commented on 2010-11-17 16:48

@Det: My point was that chrome no longer uses an external flash plugin, it does it all internally. I understand what optdepends are for.

Anonymous comment on 2010-11-16 19:28

pkgver=9.0.570.1 >> pkgver=9.0.576.0 and finish (save PKGBUILD and install)

Det commented on 2010-11-16 18:58

Sure? Gecko-mediaplayer is not needed either but they can stay there as a reminder, if x-demon prefers it that way (or mention that through the .install file, if he would like to be so pedantic about that but still include the reminder).

scio commented on 2010-11-16 17:50

@Det: Yes I agree, but I was just stating something that people might not have noticed. I would be all for removing the version number since it doesn't actually pull a specific version.

Also shouldn't flashplugin be removed from optdepends?

scio commented on 2010-11-16 17:49

@Det: Yes I agree, but I was just stating something that people might not have noticed. I would be all for removing the version number since it doesn't actually pull a specific version.

scio commented on 2010-11-16 17:49

@Det: Yes I agree, but I was just stating something that people might not have noticed. I would be all for removing the version number since it doesn't actually pull a specific version.

scio commented on 2010-11-16 17:46

@Det: Yes I agree, but I was just stating something that people might not have noticed. I would be all for removing the version number since it doesn't actually pull a specific version.

Det commented on 2010-11-16 13:10

That doesn't mean it's not out-of-date. The pkgver is the one that defines whether this thing is out of date or not - 'no matter what'.

Otherwise the pkgver shouldn't even follow the actual release numbers but instead be something like "LATEST".

scio commented on 2010-11-15 20:45

For anyone marking this as out of date, just re-install it to get the latest version. The PKGBUILD just pulls the latest dev build no matter what number is on the package.

Det commented on 2010-11-06 22:41

_Thanks!_

Anonymous comment on 2010-11-06 09:59

yeah. working on it right now.

Det commented on 2010-11-06 09:43

@x-demon, could you update your other Chromes too?

master commented on 2010-11-01 19:31

I made a package fixing Google Chrome's crash with jpeg files. Check it out: http://aur.archlinux.org/packages.php?ID=42507

Det commented on 2010-10-30 15:04

o_O What about the "bug buddy" thing? That seemed like a good idea.

t3ddy commented on 2010-10-30 12:13

@x-demon
I think should remove the args.patch since everybody can enable flags from about:flags

Det commented on 2010-10-27 14:25

@x-demon, just click the *Show all 100 comments* button on the lower left corner.

master commented on 2010-10-27 08:23

@x-demon here it is:

To fix the libjpeg6 related crash when posting JPEGs:
1. Obtain a libpixbufloader-jpeg.so that links to libjpeg6 (I stole mine from Ubuntu Karmic's GTK package)
2. Put it in /opt/google/chrome/
3. Run: sudo gdk-pixbuf-query-loaders > /opt/google/chrome/loaders.cache
4. Open the loaders.cache and change the line referring to libpixbufloader-jpeg.so to point to the one you put in /opt/google/chrome/
5. Open the google-chrome script and add somewhere before the exec: export GDK_PIXBUF_MODULE_FILE=/opt/google/chrome/loaders.cache

Both amd64 & i386 versions of Ubuntu's libgtk are here: http://packages.ubuntu.com/karmic/libgtk2.0-0

It works for me.

Anonymous comment on 2010-10-26 20:59

probably. But looks like AUR lost comments again...

master commented on 2010-10-26 20:54

rossy's libpixbufloader hack definitely fixes jpeg releated crash! Shouldn't we include it in PKGBUILD?

Det commented on 2010-10-24 16:48

I dunno why should any of them be enabled by default. It'd be better to let the user decide and at most just have a comment about these flags through the .install file. You also get the wrong impression of Chrome by using unfinished features without ever knowing about it.

t3ddy commented on 2010-10-24 09:47

Yes, I've read and tried, after posting the message before :D
It works.

@x-demon
since there are lot of flags that create problems, I think that in args.patch you should put only: "--enable-webgl --enable-accelerated-2d-canvas --enable-nacl" to have less problems possible

Det commented on 2010-10-23 20:51

Did you read their testing experiences? Most report that "--disable-accelerated-compositing" works.

t3ddy commented on 2010-10-23 19:13

@Det
It's a known bug: http://code.google.com/p/chromium/issues/detail?id=57741

Det commented on 2010-10-23 11:17

Did you clear your cache or... something :D? I dunno what else could be wrong. You could also try moving out the settings ($home/.chromium) (e.g. rename the folder ".chromium" to ".chromiuma") to have a vanilla "Dev Channel's Chrome".

t3ddy commented on 2010-10-23 11:04

@Det
You're lucky, I see html5 videos pink :S

Det commented on 2010-10-23 10:06

Heh, I'm starting to like you :).

But the "update text" says "Update: Google Chrome Frame, 7.0.517.43, has been release to stable and beta channels."

The latest stable "Google Chrome" is 7.0.517.41.

Det commented on 2010-10-23 09:36

Uh, well that was fast. But why didn't you update google-chrome/-beta on the same go?

Det commented on 2010-10-23 09:36

Uh, well that was fast. But why didn't you udpate google-chrome/-beta on the same go?

Det commented on 2010-10-23 08:50

Really? By removing it it seems I can play html5 videos just fine (tried youtube and the xiph.org's video).

Also a simple Googling showed how "--enable-gpu-plugin" and "--enable-click-to-play" (used here) are deprecated/obsolete, while "--enable-vertical-tabs" isn't yet implemented on linux - so at least the first two should be removed.

t3ddy commented on 2010-10-22 21:09

It's enabled by default (by the maintainer), look at args.patch.
Anyway removing that flag will let you see html5 videos, in a bright pink.

t3ddy commented on 2010-10-22 21:08

It's enabled by default, look at arg.patch.
Anyway removing that flag will let you see html5 videos, in a bright pink.

t3ddy commented on 2010-10-22 21:07

It's enabled by default, look at arg.patch.
Anyway removing that flag will let you see html5 videos, in a bright pink.

Det commented on 2010-10-22 17:37

Remove it? I didn't have it in the first place. Or do you mean it's enabled by default? In which case you mean "--disable-video-layering"?

t3ddy commented on 2010-10-22 16:31

you should remove --enable-video-layering flag since it is not yet completly supported

Det commented on 2010-10-16 10:31

For some reason I have no video in html5 videos because of this "no such file or directory" plugin error (http://aur.pastebin.com/pGJp3Ldg). On the other hand chromium-browser-bin with chromium-codecs-ffmpeg-nonfree-bin works fine.

Weren't the Ffmpeg codecs supposed to come with Google Chrome?

E: Hmm, well it seems there's a few bug reports about Google Chrome and html5 already done. So I guess this is a know issue atm.

Det commented on 2010-10-16 10:28

For some reason I have no video in html5 videos because of this "no such file or directory" plugin error (http://aur.pastebin.com/pGJp3Ldg). On the other hand chromium-browser-bin with chromium-codecs-ffmpeg-nonfree-bin works fine.

Weren't the Ffmpeg codecs supposed to come with Google Chrome?

t3ddy commented on 2010-10-13 13:05

Does this video (http://xiph.org/video/vid1.shtml) works for you?

Anonymous comment on 2010-10-08 07:43

Probably a better solution would be to build a libpixbufloader-jpeg.so statically linked to libjpeg8. That way programs launched by Chrome (like when you download and run an executable) won't crash.

Det commented on 2010-10-07 22:48

Ah, of course. Since it's for libjpeg6 to which this whole thing is built for. I somehow got the idea that you just needed the file itself in that 'hack'.

OttoA commented on 2010-10-07 19:37

@Det thats the recent libpixbufloader-jpeg.so - for chrome to run properly you need an older one

Det commented on 2010-10-07 18:38

Hmm, good to know. But gdk-pixbuf2's ([testing]) libpixbufloader-jpeg.so isn't enough?

Anonymous comment on 2010-10-07 14:58

To fix the libjpeg6 related crash when posting JPEGs:
1. Obtain a libpixbufloader-jpeg.so that links to libjpeg6 (I stole mine from Ubuntu Karmic's GTK package)
2. Put it in /opt/google/chrome/
3. Run: sudo gdk-pixbuf-query-loaders>/opt/google/chrome/loaders.cache
4. Open the loaders.cache and change the line referring to libpixbufloader-jpeg.so to point to the one you put in /opt/google/chrome/
5. Open the google-chrome script and add somewhere before the exec: export GDK_PIXBUF_MODULE_FILE=/opt/google/chrome/loaders.cache

I don't know how I would go about writing a patch or PKGBUILD that does this, but would it work? It works for me.

Det commented on 2010-10-06 14:47

See here: http://code.google.com/p/chromium/issues/detail?id=54932#c5

haagch commented on 2010-09-30 22:31

Are there any news about --enable-gpu-rendering?
I have not seen it work on any of the following: intel 945gme with xf86-video-intel, radeon hd 4670 with either catalyst or xf86-video-ati-git...

Anonymous comment on 2010-09-30 09:30

update please

beej commented on 2010-09-29 01:07

It is a little funky to have the package version out of sync with the chrome version. I've been modding the PKGBUILD by hand when a new rev comes out.

I certainly never minded the frequent updates to the package--it let me know when a new dev channel build was out.

Just my $0.02.

Det commented on 2010-09-17 12:26

@Zom, if you want to do that with your package, I won't step in your way but I am not going to redownload a package that has specific release versioning again every time a new version comes up.

You might as well not change the pkgrel at all, if you think your ideology is correct - which it is not.

Anonymous comment on 2010-09-17 10:58

who?

Zom commented on 2010-09-17 10:55

Could you *please* stop marking this as out of date? The PKGBUILD fetches the latest version, so just reinstall whenever you want an update.

Flag it out of date when either it's missing certain dependencies or when dependencies become obsolete.

OttoA commented on 2010-09-13 10:11

If you have issues with flash sound (too loud / clipping noise), disable the internal flash plugin with about:plugins -> click on details link -> disable the plugin under "/opt/google/chrome"

OttoA commented on 2010-09-09 16:00

@mccatec: yes this is a known issue. Chromium-browser-svn builds with libjpeg8 while this needs libjpeg6. Gnome (and thereby the open file dialog) is built with libjpeg8, so it crashes

Anonymous comment on 2010-09-09 14:48

The browser still crashes when I try to upload a jpg file from local disk. Does this problem have something to do with libjpeg6? Since the package ``chromium-browser-svn'', which relies on libjpeg instead of libjpeg6, does not crash at the same situation.
Thx

trizen commented on 2010-09-09 13:53

Hmm... After update to v7.0.517.0 CPU is 100% all the time...

Anonymous comment on 2010-09-09 13:09

patch is also fixed.

Anonymous comment on 2010-09-09 13:02

okay, so now i can say for sure that GPU rendering completely broken under linux with any GPU card.

Anonymous comment on 2010-09-09 12:55

@x-demon
on x86,with intel video card and the driver 'xf86-video-intel',enable gpu rendering causes black window,the rest args seems work well so far.
thanks for ur work!

Anonymous comment on 2010-09-09 08:46

Yes, patching failed because new version released
Comment this lines

msg "Patching launcher"
cd $pkgdir/opt/google/chrome
patch -p0 < $srcdir/args.patch || return 1
msg2 "Done patching!"

in pkgbuild file & restart installing
works.

Anonymous comment on 2010-09-09 05:51

==> Patching launcher
patching file google-chrome
Hunk #1 FAILED at 30.
1 out of 1 hunk FAILED -- saving rejects to file google-chrome.rej
Преждевременный выход...
==> ERROR: Makepkg was unable to build google-chrome-dev.


version 517 package build error during patching :(

Det commented on 2010-09-08 16:26

Says a person who got his name in both a 'contributor' _and_ a 'maintainer' line. (You can) Read the reason why that's not-so sane in the google-chrome-beta comment section.

Anonymous comment on 2010-09-06 13:21

@EL_Caballero
read the f*cking manual

Anonymous comment on 2010-09-06 12:44

It has some requisited missed. For example, patch, bison and flex

OttoA commented on 2010-09-05 18:58

x86, nvidia, xinerama user here: the switch --enable-gpu-rendering makes the window black (same as amirs) and --enable-video-layering makes html5 videos go black

OttoA commented on 2010-09-05 18:47

html5 video on youtube doesn't work anymore (black window), same issue as amirs with --enable-gpu-rendering

Anonymous comment on 2010-09-04 15:43

Same as amirs here.

Anonymous comment on 2010-09-03 12:57

i have glitches with xf86-video-ati-git, by the way.

amirs commented on 2010-09-03 12:42

on x86_64 with nvidia proprietary, --enable-gpu-rendering renders as black window. all the rest seems to work and seems to be much faster.

Anonymous comment on 2010-09-03 12:10

should i include new args options?
--enable-gpu-plugin --enable-accelerated-compositing --enable-click-to-play --enable-gpu-rendering --enable-video-layering --enable-webgl --enable-accelerated-2d-canvas --enable-nacl --enable-vertical-tabs
please test chrome with this arguments (run google-chrome from cli with that args) and report if that works correctly.
New GPU option gives chrome insane speed at rendering
(i need reports from people with nvidia, intel and ati videocards, with closed and open-source drivers)
thanks!

dsadcsadsadacasd commented on 2010-08-27 05:59

Re-installing package install latest version.

Anonymous comment on 2010-08-13 16:49

updated.

Anonymous comment on 2010-08-13 16:11

Today's update fixes the pdf plugin and some other bugs. PKGBUILD doesn't check md5 sum, so you can just rebuild and get newest version.

Anonymous comment on 2010-07-30 07:25

pdf currently broken in Mac and Linux

Anonymous comment on 2010-07-30 06:10

Hmm... lost PDF plugin support in the latest update.

Anonymous comment on 2010-07-29 16:35

i was at vacation in spain, so sorry for delay. Updated.

laloch commented on 2010-07-29 13:57

@burgua: URL in the PKGBUILD is http://dl.google.com/linux/direct/google-chrome-${_chanswitch}_current_$ARCH.deb, so you can always get the latest version by reinstalling the package. You should also modify pkgver variable, but that's not mandatory.

Anonymous comment on 2010-07-28 20:13

Personally, I don't mind it being out-of-date (for the time being) because I consider the pdf plugin more important than the changes in the recent updates.

Anonymous comment on 2010-07-28 05:45

update please

Anonymous comment on 2010-07-09 18:18

pdf works for me.
BTW, since it's a -dev version, i decided to enable new wrench menu and chrome apps, since that functions stable for me. Let me know is there any problem with that.

Anonymous comment on 2010-07-05 12:38

The new PDF plugin (which appears in 6.0.453.1 for Linux) doesn't appear to be built, I've got nothing in about:plugins.

Anonymous comment on 2010-07-04 23:21

Got build 6.0.453.1 today. Pinned tabs disappear when you close them now... what happened? :(

nicedream commented on 2010-07-01 01:01

You should add the google-chrome package as a conflict, since it is not possible to install both at the same time (they both install a lot of the same files).

Or, ideally, why not move this package to its own directory? That way a user can run the stable and the dev package at the same time.

laloch commented on 2010-06-25 21:21

Current linux version is 6.0.437.3. Build 447.0 id Mac and Windows only.

Anonymous comment on 2010-06-20 11:49

Unless you need specific features of this official Google dev channel version, chromium-dev is a better PKGBUILD. It builds from source, so it takes some time at each upgrade, but the maintainer seems to update it fairly quickly.

t3ddy commented on 2010-05-27 17:00

you're still creating a directory called google-chrome.png...

mortzu commented on 2010-05-25 15:39

please update the url of the 64bit package to the new one

Anonymous comment on 2010-05-22 08:56

Revelation60, indeed. the new location is http://dl.google.com/linux/direct/google-chrome-unstable_current_amd64.deb so something along the lines of t3ddy's PKGBUILD should work. I quoted your comment because sovok_kpss seemed dismissive and i thought i had missed something.

Revelation60 commented on 2010-05-22 08:34

jai, that link used to work. Unfortunately, it doesn't work anymore.

t3ddy commented on 2010-05-22 07:12

what about a pkgbuild like this: http://aur.pastebin.com/YGCni0Ka ?

Anonymous comment on 2010-05-22 04:48

Err, which comment are you referring to?
"Ramblurr, this has been said a couple of times now. What you can do is edit the PKGBUILD to change the release to 47050" <--- This?
I'm having problems with the the latest PKGBUILD with _revision set to 47050.

sovok_kpss commented on 2010-05-22 04:33

Please, read the previous comments

Anonymous comment on 2010-05-21 15:15

Broken on x86_64
ar: google-chrome-unstable_6.0.401.1-r47050_amd64.deb: No such file or directory

Anonymous comment on 2010-05-20 12:18

yes, google chrome linked against that libraries.

nbvcxz commented on 2010-05-20 12:17

Is it really necessary to have libjpeg6, libpng12 (instead of libjpeg libpng)?

Anonymous comment on 2010-05-20 11:43

i dunno when google will bring back old URL, but i believe i temporary fixed problem with x86_64

Anonymous comment on 2010-05-20 11:40

because it was a suggestion, and not from one user.

sjakub commented on 2010-05-20 11:37

This package looks horrible. Why do you use wget instead of standard method?

Revelation60 commented on 2010-05-20 08:05

Ramblurr, this has been said a couple of times now. What you can do is edit the PKGBUILD to change the release to 47050.

Anonymous comment on 2010-05-20 03:22

this package is broken. the deb package it downloads no longer exists:


==> Downloading Google Chrome Distro
--2010-05-19 23:21:58-- http://dl.google.com/linux/deb/pool/main/g/google-chrome-unstable/google-chrome-unstable_6.0.401.1-r47049_amd64.deb
Resolving dl.google.com... 72.14.204.136, 72.14.204.190, 72.14.204.91, ...
Connecting to dl.google.com|72.14.204.136|:80... connected.
HTTP request sent, awaiting response... 404 Not Found
2010-05-19 23:21:59 ERROR 404: Not Found.

-> Done downloading!
==> Extracting Google Chrome
ar: google-chrome-unstable_6.0.401.1-r47049_amd64.deb: No such file or directory
==> ERROR: Build Failed.
Aborting...

sud_crow commented on 2010-05-20 00:56

Missing icon for KDE 4.4 too.
Regards.

Anonymous comment on 2010-05-19 20:48

Could you update dependencies? I have a newer version from libjpeg6 and libpng12.

pacman -Si libjpeg
Repository : extra
Name : libjpeg
Version : 8.0.1-1

yetAnotherZero commented on 2010-05-19 17:17

the download link is giving a 404 for a couple days now...

t3ddy commented on 2010-05-17 19:38

why a directory named google-chrome.png??
shouldn't it be like:

for i in 16 32 48 256;
do
install -d ${pkgdir}/usr/share/icons/hicolor/${i}x${i}/apps
cp -v ${pkgdir}/opt/google/chrome/product_logo_${i}.png ${pkgdir}/usr/share/icons/hicolor/${i}x${i}/apps/google-chrome.png
done

?

t3ddy commented on 2010-05-17 19:22

why a directory named google-chrome.png??

Anonymous comment on 2010-05-17 15:58

Ctrl+B does not works only for me?
Crashing when select a *.jpg file from filemanager - for me too?
And for 6.0... missing icon in the main menu (KDE 4.4)

Anonymous comment on 2010-05-17 01:14

The temp link provided by sovok_kpss for x86_64 below actually works, but since, for some reason, _revision variable (well, at least, it should be a variable) is hard-coded in PKGBUILD, you just need to edit it and replace 47049, which apparently still works for the x86 version, with 47050, to get it to fetch and build the package on amd64.

sovok_kpss commented on 2010-05-16 06:07

amd64 temp link:
http://dl.google.com/linux/deb/pool/main/g/google-chrome-unstable/google-chrome-unstable_6.0.401.1-r47050_amd64.deb

Anonymous comment on 2010-05-15 15:07

you can temporarily change the download link to:
http://dl.google.com/linux/deb/pool/main/g/google-chrome-unstable/google-chrome-unstable_6.0.401.1-r47049_i386.deb

and of course some small changes to PKGBUILD should be made.

Anonymous comment on 2010-05-15 15:01

i know about 404. There is _no_ file on _google_ server. Stop that flood.

Anonymous comment on 2010-05-15 14:38

and +1 to mvorozhtsov

Anonymous comment on 2010-05-15 10:25

404 Not Found?

Anonymous comment on 2010-05-15 10:14

looks like problems on google side - links from their site also have 404... btw, i updated pkgbuild.

Anonymous comment on 2010-05-15 09:30

error 404 when downloading http://dl.google.com/linux/direct/google-chrome-unstable_current_i386.deb

Anonymous comment on 2010-05-14 06:26

There is an error in icons installation code:
$ pacman -Ql google-chrome-dev | grep hicolor | tail -3
google-chrome-dev /usr/share/icons/hicolor/48x/apps/
google-chrome-dev /usr/share/icons/hicolor/48x/apps/google-chrome.png/
google-chrome-dev /usr/share/icons/hicolor/48x/apps/google-chrome.png/product_logo_48.png

${i}x${x} should be replaced with ${i}x${i}. google-chrome-beta PKGBUILD needs to be fixed too.

Anarconda commented on 2010-04-25 08:17

[amarcord@Gaia ~]$ /opt/google/chrome/chrome
/opt/google/chrome/chrome: error while loading shared libraries: libnss3.so.1d: cannot open shared object file: No such file or directory

but:

[root@Gaia google-chrome-dev]# ls -l /opt/google/chrome/lib/libnss3.so.1d
lrwxrwxrwx 1 root root 19 abr 25 10:03 /opt/google/chrome/lib/libnss3.so.1d -> /usr/lib/libnss3.so

What's wrong? TIA.

Anonymous comment on 2010-04-25 07:57

looks like bug in aurploader :) updated.

Anonymous comment on 2010-04-25 07:54

What is stopping this from being updated to 5.0.375.19 ?

valandil commented on 2010-04-14 01:41

If Chrome v.5.0.371.0 crashes at boot for you, consider moving your Sync Data directory (~/.config/google-chrome/Default/Sync Data) to somewhere else and reboot it again. Chances are it'll work again in no time.

valandil commented on 2010-04-12 15:37

Hello! Just updated my system with [testing] repo enabled, and Chrome crashes at boot systematically. At first, he was looking for libraries that did not exist, so I used:

sudo ln -s /usr/lib/libnss3.so /usr/lib/libnss3.so.1d
sudo ln -s /usr/lib/libnssutil3.so /usr/lib/libnssutil3.so.1d
sudo ln -s /usr/lib/libsmime3.so /usr/lib/libsmime3.so.1d
sudo ln -s /usr/lib/libssl3.so /usr/lib/libssl3.so.1d
sudo ln -s /usr/lib/libplds4.so /usr/lib/libplds4.so.0d
sudo ln -s /usr/lib/libplc4.so /usr/lib/libplc4.so.0d
sudo ln -s /usr/lib/libnspr4.so /usr/lib/libnspr4.so.0d
sudo ln -s /usr/lib/libavutil.so.50 /usr/lib/libavutil.so.49

The error is now:

[5603:5624:264713217:FATAL:chrome/browser/sync/syncable/directory_backing_store.cc(456)] Check failed: SQLITE_DONE == step_result.

Happens to you too?

Anonymous comment on 2010-04-09 14:23

i updated package, added --enable-webgl and removed --no-sandbox as suggested by valandil

valandil commented on 2010-04-09 13:59

Regarding version 5.0.371.0, one may consider removing the --no-sandbox flag, as it is no longer required for WebGL.

Would improve security and stability.

liquibyte commented on 2010-04-03 00:49

Build fail

lzma: (stdin): Compressed data is corrupt
tar: Unexpected EOF in archive
tar: Unexpected EOF in archive
tar: Error is not recoverable: exiting now

Anonymous comment on 2010-04-02 17:09

The new Chrome-devs all include Flash? Has this been implemented in Arch yet? Should I get rid of Flash from aur and instead just let chrome keep it up to date for me? Or is it not bundled with Archs version yet?

Anonymous comment on 2010-03-30 14:55

5.0.360.0 crashes all the time

Anonymous comment on 2010-03-28 15:08

x-demon, change version package on 5.0.360.0

Anonymous comment on 2010-03-24 17:21

Anyone having problems with google-chrome-dev 5.0.356.2-1 and flash?