Feature #15803
Upgrade Tor Browser to 8.0a10
100%
Description
It’s not the final 8.0 but better detect now any breakage or adjustment we need to do on our side.
Subtasks
Related issues
Blocks Tails - |
Resolved | 2018-02-20 | |
Blocks Tails - |
Resolved | 2018-07-03 | |
Blocks Tails - |
Resolved | 2018-08-17 | |
Blocks Tails - |
Resolved | 2018-08-18 |
History
#1 Updated by intrigeri 2018-08-18 07:51:05
- Description updated
#2 Updated by intrigeri 2018-08-18 07:51:17
- blocks
Feature #15334: Core work 2018Q3: Foundations Team added
#3 Updated by intrigeri 2018-08-18 12:11:41
- Status changed from Confirmed to In Progress
- % Done changed from 0 to 10
- Feature Branch set to feature/15803-tor-browser-8.0a10+force-all-tests
#4 Updated by intrigeri 2018-08-18 12:11:54
- blocks
Bug #15706: Tor Browser 8 always prompts wrt. asking webpages in English added
#5 Updated by intrigeri 2018-08-18 12:34:57
- blocks
Bug #15801: The Tor Browser 8.0a9 icon (overview, taskbar) is uglier than it used to be added
#6 Updated by intrigeri 2018-08-18 13:44:17
- blocks
Bug #15805: Update doc for Tor Browser 8 added
#7 Updated by intrigeri 2018-08-18 16:25:45
Note to myself:
16:01:43 The following supported browser locales lack descriptions in /usr/share/tails/browser-localization/descriptions: ca ga-IE id is nb-NO
16:01:43 E: config/chroot_local-hooks/11-localize_browser failed (exit non-zero). You should check for errors.
#8 Updated by intrigeri 2018-08-19 15:58:47
- Assignee changed from intrigeri to segfault
- % Done changed from 10 to 50
- Estimated time set to 1 h
- QA Check set to Ready for QA
Note: it fixes Bug #15777 for NoScript :)
#9 Updated by segfault 2018-08-23 10:55:13
I reviewed the code and have the following notes:
97556aaed14dca855269709ae753ebcf02cf3bf8:
- The release doc doesn’t say what to do if the glibc version changed
- Why don’t we use Tor Browser’s abicheck in chroot_local-hooks/10-tbb to determine whether to use the system libc?
I want to build and test the branch, but can’t do that right now.
#10 Updated by intrigeri 2018-08-24 05:32:07
> I reviewed the code
Thank you!
> 97556aaed14dca855269709ae753ebcf02cf3bf8:
> * The release doc doesn’t say what to do if the glibc version changed
Right. Fixed in commit:efcbcb65322860a471c5068f32c92d79eb446663.
> * Why don’t we use Tor Browser’s abicheck in chroot_local-hooks/10-tbb to determine whether to use the system libc?
I’m not confident a check done in 10-tbb
will always be reliable: sometimes we install/upgrade stuff in later hooks. So the only way to use that abicheck
program would be at runtime and then it gets a bit too complicated IMO. Now, I think it’s too early to tell what will work for us, so I’d rather do the simpler thing now (as my branch does) and if this causes maintenance trouble in the future, it’ll still be time to change our mind and use their abicheck
. Fair enough?
#11 Updated by segfault 2018-08-24 11:03:18
intrigeri wrote:
> Right. Fixed in commit:efcbcb65322860a471c5068f32c92d79eb446663.
Great
> > * Why don’t we use Tor Browser’s abicheck in chroot_local-hooks/10-tbb to determine whether to use the system libc?
>
> I’m not confident a check done in 10-tbb
will always be reliable: sometimes we install/upgrade stuff in later hooks.
I see. Maybe we could do it in a separate 99-
hook.
> So the only way to use that abicheck
program would be at runtime and then it gets a bit too complicated IMO. Now, I think it’s too early to tell what will work for us, so I’d rather do the simpler thing now (as my branch does) and if this causes maintenance trouble in the future, it’ll still be time to change our mind and use their abicheck
. Fair enough?
Sure
#12 Updated by segfault 2018-08-24 11:26:46
- Assignee changed from segfault to intrigeri
- QA Check changed from Ready for QA to Pass
> I want to build and test the branch, but can’t do that right now.
Done, everything looking good.
> Note: it fixes Bug #15777 for NoScript :)
Indeed, that’s great! HTTPS Everywhere is still affected, but I think that is less of a problem.
#13 Updated by intrigeri 2018-08-25 07:37:13
- Assignee changed from intrigeri to segfault
- % Done changed from 50 to 90
- QA Check changed from Pass to Ready for QA
Great, thanks! But before I merge the branch I need your QA on Bug #15706 and Bug #15801. Did you do them at the same time already?
#14 Updated by segfault 2018-08-25 11:12:22
- QA Check changed from Ready for QA to Pass
intrigeri wrote:
> Great, thanks! But before I merge the branch I need your QA on Bug #15706 and Bug #15801. Did you do them at the same time already?
I tested Bug #15801 at the same time already, and now also tested Bug #15706
#15 Updated by intrigeri 2018-08-26 05:51:27
> I tested Bug #15801 at the same time already, and now also tested Bug #15706
Thanks a lot!
#16 Updated by intrigeri 2018-08-26 05:56:29
- Status changed from In Progress to Fix committed
- Assignee deleted (
segfault) - % Done changed from 90 to 100
#17 Updated by sajolida 2018-09-02 17:10:36
I don’t dare opening a new ticket for that and assume that you’re on the case already but in a recent testing image (31c1a8c646) I still have JavaScript disabled by default.
I showed you that some days ago and I understand that it was related to this issue.
#18 Updated by intrigeri 2018-09-02 20:44:31
> I don’t dare opening a new ticket for that and assume that you’re on the case already but in a recent testing image (31c1a8c646) I still have JavaScript disabled by default.
I’m not on the case already and without a ticket, this totally slipped off my radar. In the future please file tickets for such critical issues (you dare filing tickets for much less critical ones, which is good, and I encourage you to do the same for potential release blockers :)
I’ve filed Bug #15905 about this.
#19 Updated by intrigeri 2018-09-05 16:15:42
- Status changed from Fix committed to Resolved