WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
190241
[WPE][GTK] run-minibrowser improperly creates webkit-flatpak environment
https://bugs.webkit.org/show_bug.cgi?id=190241
Summary
[WPE][GTK] run-minibrowser improperly creates webkit-flatpak environment
Michael Catanzaro
Reported
2018-10-03 02:42:07 PDT
I'm using a jhbuild environment, NOT webkit-flatpak. Every time I run run-minibrowser it prints this: $ run-minibrowser --gtk You need to install flatpak-builder >= 0.10.0 to be able to use the '/home/mcatanzaro/Projects/WebKit/Tools/Scripts/webkit-flatpak' script. You can find some informations about how to install it for your distribution at: *
http://flatpak.org/
And it creates GTK/ and UserFlatpak/ subdirectories under my build directory. If I delete them, they are created again the next time I run run-minibrowser.
Attachments
Patch
(2.59 KB, patch)
2018-12-09 15:09 PST
,
Michael Catanzaro
no flags
Details
Formatted Diff
Diff
Patch for landing
(1.04 KB, patch)
2018-12-10 16:01 PST
,
Adrian Perez
no flags
Details
Formatted Diff
Diff
Patch for landing
(1.78 KB, patch)
2018-12-10 16:05 PST
,
Adrian Perez
mcatanzaro
: commit-queue-
Details
Formatted Diff
Diff
Show Obsolete
(2)
View All
Add attachment
proposed patch, testcase, etc.
Thibault Saunier
Comment 1
2018-10-03 04:06:36 PDT
You somehow have a trailling flatpak env?
Michael Catanzaro
Comment 2
2018-10-03 07:17:16 PDT
No, I delete it before running run-minibrowser. run-minibrowser is creating it.
Michael Catanzaro
Comment 3
2018-11-02 09:24:19 PDT
Thibault, can you look into this please?
Michael Catanzaro
Comment 4
2018-12-09 15:06:33 PST
Problem is flatpakutils.py unconditionally creates the directories, but should do so only when called with --available.
Michael Catanzaro
Comment 5
2018-12-09 15:09:39 PST
Created
attachment 356930
[details]
Patch
Ross Kirsling
Comment 6
2018-12-09 15:13:16 PST
Nice! I noticed this happening for WinCairo too but hadn't looked into why.
WebKit Commit Bot
Comment 7
2018-12-10 06:58:39 PST
Comment on
attachment 356930
[details]
Patch Clearing flags on attachment: 356930 Committed
r239028
: <
https://trac.webkit.org/changeset/239028
>
WebKit Commit Bot
Comment 8
2018-12-10 06:58:41 PST
All reviewed patches have been landed. Closing bug.
Radar WebKit Bug Importer
Comment 9
2018-12-10 06:59:26 PST
<
rdar://problem/46594304
>
Adrian Perez
Comment 10
2018-12-10 16:00:15 PST
This broke all the Flatpak machinery, any tool that tries to use the utility module will complain with: Traceback (most recent call last): File "/home/aperez/devel/WebKit-ContentBlocking/Tools/Scripts/update-webkitwpe-flatpak", line 28, in <module> WebkitFlatpak.load_from_args(["--wpe", "--update"] + sys.argv[1:]).run() File "/home/aperez/devel/WebKit-ContentBlocking/Tools/flatpak/flatpakutils.py", line 761, in run if not self.clean_args(): File "/home/aperez/devel/WebKit-ContentBlocking/Tools/flatpak/flatpakutils.py", line 574, in clean_args _log.debug("Using flatpak user dir: %s" % os.environ["FLATPAK_USER_DIR"]) File "/home/aperez/.virtualenvs/webkit/lib/python2.7/UserDict.py", line 40, in __getitem__ raise KeyError(key) KeyError: 'FLATPAK_USER_DIR'
Adrian Perez
Comment 11
2018-12-10 16:01:20 PST
Created
attachment 357010
[details]
Patch for landing
Adrian Perez
Comment 12
2018-12-10 16:05:02 PST
Created
attachment 357011
[details]
Patch for landing
Michael Catanzaro
Comment 13
2018-12-10 16:20:27 PST
Comment on
attachment 357011
[details]
Patch for landing This will bring back the original bug. The flatpak directory will be created if flatpak is not in use, and even for ports that don't use flatpak. If it's not enough to move this up to the top of the function, then we'll need to find another way.
Michael Catanzaro
Comment 14
2018-12-10 16:23:23 PST
Yeah it looks like this is only used in clean_args, so I would just move it up to the top of clean_args.
Michael Catanzaro
Comment 15
2018-12-15 09:55:39 PST
Ping, isn't webkit-flatpak unusable until this lands?
Michael Catanzaro
Comment 16
2019-01-01 10:26:24 PST
I guess I'm going to close this, since the original issue is fixed. I know this has broken the flatpak environment, but the fix is quite trivial, as explained in
comment #14
. Anyone who is actually using the flatpak environment should feel free to test and land that follow-up fix. In the meantime, it's clear nobody besides Adrian has attempted to use the flatpak environment since December 10, at least....
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug