DFPI breaks stackblitz.com
Categories
(Core :: Privacy: Anti-Tracking, defect, P2)
Tracking
()
Tracking | Status | |
---|---|---|
firefox-esr78 | --- | unaffected |
firefox-esr91 | --- | wontfix |
firefox92 | --- | wontfix |
firefox93 | --- | wontfix |
firefox94 | --- | verified |
People
(Reporter: emilio, Assigned: twisniewski)
References
(Blocks 1 open bug, Regression)
Details
(Keywords: regression)
Attachments
(1 file)
STR:
ER:
- A green square.
AR:
- Right tab doesn't load, there's a cross-origin error there.
Comment 1•4 years ago
|
||
This breakage can be fixed by changing cookiebehavior
to 4, so this is only related to dFPI.
Comment 2•4 years ago
|
||
It looks like the right tab is put under a third-party iframe https://rxjs-yo2jz5.stackblitz.io
and its storage is partitioned. And it needs storage access to show the content properly.
I think the proper solution is to ask the website to use the StorageAccessAPI.
Steve, do you know if we have contacted them about this issue? Thanks.
Comment 3•4 years ago
|
||
As far as I know we have not. We'll have to see if stpeter or webcompat folks have a contact.
Comment 4•4 years ago
|
||
Filed an issue on their GitHub repo: https://github.com/stackblitz/core/issues/1374
Updated•4 years ago
|
Assignee | ||
Comment 7•3 years ago
|
||
Updated•3 years ago
|
Comment 9•3 years ago
|
||
bugherder |
Updated•3 years ago
|
Comment 10•3 years ago
|
||
Verified as fixed on the latest Nightly 95.0a1 and on Firefox 94 beta 1 (having the pref network.cookie.cookieBehavior set to 5) - the green square is properly displayed on stackblitz.com.
Tested on Windows 10 x64, Ubuntu 20.04 x64, and macOS Big Sur 11.6.
Updated•3 years ago
|
Comment 11•3 years ago
|
||
StackBlitz staffer here. Sorry if this is not the right venue, but is there a way for us to test in Firefox without this shim?
We have features that are breaking because of it (e.g. https://github.com/stackblitz/core/issues/1679 and a bunch of duplicate reports), because they rely on Service Workers, and we can't test fixes in Firefox because the shim seems to apply no matter what.
I tried setting a few about:config
settings, but am not seeing a difference. I’m still getting a warning in the Console with
Service Workers are disabled by Firefox on this page to prevent breakage. See https://bugzilla.mozilla.org/show_bug.cgi?id=1668408 for details.
The about:config
options I tried (in Firefox Nightly and stable) were:
extensions.webcompat.enable_shims = false
extensions.webcompat.perform_injections = false
extensions.webcompat.perform_ua_overrides = false
Comment 12•3 years ago
|
||
(In reply to Florens Verschelde :fvsch from comment #11)
StackBlitz staffer here. Sorry if this is not the right venue, but is there a way for us to test in Firefox without this shim?
We have features that are breaking because of it (e.g. https://github.com/stackblitz/core/issues/1679 and a bunch of duplicate reports), because they rely on Service Workers, and we can't test fixes in Firefox because the shim seems to apply no matter what.
I tried setting a few
about:config
settings, but am not seeing a difference. I’m still getting a warning in the Console withService Workers are disabled by Firefox on this page to prevent breakage. See https://bugzilla.mozilla.org/show_bug.cgi?id=1668408 for details.
The
about:config
options I tried (in Firefox Nightly and stable) were:
extensions.webcompat.enable_shims = false
extensions.webcompat.perform_injections = false
extensions.webcompat.perform_ua_overrides = false
Navigate to about:compat
and click "Disable" for StackBlitz. That should disable the shim for the current browser session.
Comment 13•3 years ago
|
||
Thanks a bunch :)
Comment 14•2 years ago
|
||
Do we still need this shim, now that we support partitioned service workers? I've tested the original STR on Nightly with the shim disabled via about:compat and it seems to load just fine.
Comment 15•2 years ago
|
||
Yes, we have enabled the partitioned service workers since Fx 105. And I can confirm that it fixes this issue on stackblitz with the shim disabled. So, I think we can remove the shim.
Comment 16•2 years ago
|
||
Awesome! I'll remove it in the bug 1795692.
Description
•