RESOLVED FIXED 55109
REGRESSION(r79396): gmail is not rendered correctly on TOT
https://bugs.webkit.org/show_bug.cgi?id=55109
Summary REGRESSION(r79396): gmail is not rendered correctly on TOT
Ryosuke Niwa
Reported 2011-02-23 18:16:04 PST
Gmail's inbox has extra padding in each line on TOT but it doesn't have such a padding on Safari / Chrome. In fact, this is a regression happened in the last 48 hours.
Attachments
correct (Safari) (50.04 KB, image/png)
2011-02-23 18:17 PST, Ryosuke Niwa
no flags
incorrect (WebKit TOT) (86.16 KB, image/png)
2011-02-23 18:18 PST, Ryosuke Niwa
no flags
fixes the bug (4.30 KB, patch)
2011-02-23 20:35 PST, Ryosuke Niwa
no flags
incorrect rendering of thread view (101.59 KB, image/png)
2011-02-25 21:11 PST, Ryosuke Niwa
no flags
Ryosuke Niwa
Comment 1 2011-02-23 18:17:40 PST
Created attachment 83595 [details] correct (Safari)
Ryosuke Niwa
Comment 2 2011-02-23 18:18:04 PST
Created attachment 83596 [details] incorrect (WebKit TOT)
Alexey Proskuryakov
Comment 3 2011-02-23 18:52:11 PST
Ryosuke Niwa
Comment 4 2011-02-23 19:01:08 PST
I suspect that this is due to http://trac.webkit.org/changeset/79396/.
Ryosuke Niwa
Comment 5 2011-02-23 19:10:54 PST
Confirmed. This is a regression due to r79396. Perhaps we can't drop the language name given. I'm afraid that numerous other websites might be broken due to this change.
Ryosuke Niwa
Comment 6 2011-02-23 19:14:17 PST
Ojan Vafai
Comment 7 2011-02-23 20:05:07 PST
Gmail uses a Closure's UA parsing code (http://www.google.com/codesearch/p?hl=en#epIciakqvFc/trunk/closure/goog/useragent/useragent.js&q=userAgent%20package:http://closure-library%5C.googlecode%5C.com&sa=N&cd=1&ct=rc), so there are likely a lot of Google, and possibly non-Google, products that will have problems here. Although, I'm not at all sure what's going wrong here. In either case, we can certainly get the library fixed. It's just a question of how long all the sites that use the code have this problem. I'll see if I can figure out what problem this is actually causing.
Ojan Vafai
Comment 8 2011-02-23 20:06:38 PST
Actually, sorry, looking at that code, there's no way that's the culprit. Maybe it's server-side UA detection.
Ojan Vafai
Comment 9 2011-02-23 20:14:56 PST
Looks like the padding on the TDs in each row gets overridden. On tip-of-tree it doesn't.
Ryosuke Niwa
Comment 10 2011-02-23 20:35:31 PST
Created attachment 83599 [details] fixes the bug
Alexey Proskuryakov
Comment 11 2011-02-23 20:39:07 PST
Comment on attachment 83599 [details] fixes the bug Wrong bug.
Ryosuke Niwa
Comment 12 2011-02-23 20:51:15 PST
(In reply to comment #11) > (From update of attachment 83599 [details]) > Wrong bug. Oops! My bad :(
Emil A Eklund
Comment 13 2011-02-24 11:10:13 PST
Gmail uses server side user agent detection to determine which version of the stylesheet to send. I'll talk to the gmail team.
Ryosuke Niwa
Comment 14 2011-02-24 18:00:13 PST
It seems like Google is going to fix the server-side UA string parser to resolve this particular issue. But I bet there are similar regressions happening all over the place. Do we have any plans to do some sort of public announcements? A blog post would work.
Ryosuke Niwa
Comment 15 2011-02-25 21:11:54 PST
Created attachment 83922 [details] incorrect rendering of thread view
Alexey Proskuryakov
Comment 16 2011-02-28 17:16:15 PST
*** Bug 55398 has been marked as a duplicate of this bug. ***
Ryosuke Niwa
Comment 17 2011-02-28 17:46:50 PST
*** Bug 55390 has been marked as a duplicate of this bug. ***
Alexey Proskuryakov
Comment 18 2011-03-01 16:46:40 PST
> It seems like Google is going to fix the server-side UA string parser to resolve this particular issue. Did you happen to get an ETA for that?
Peter Kasting
Comment 19 2011-03-02 09:21:06 PST
(In reply to comment #18) > > It seems like Google is going to fix the server-side UA string parser to resolve this particular issue. > > Did you happen to get an ETA for that? Yes, it should be available within a week.
Peter Kasting
Comment 20 2011-03-02 09:22:08 PST
I'm going to mark this WONTFIX since the issue is server-side, and the server code has already been fixed and just needs to be pushed out.
Alexey Proskuryakov
Comment 21 2011-03-02 11:34:26 PST
Evangelism/FIXED seems more appropriate. It's not like we decided that GMail looking bad doesn't need fixing!
Jason Levine
Comment 22 2011-03-09 06:29:35 PST
(In reply to comment #19) Peter et al, is there any chance that this fix is going to be pushed out soon? This isn't a *huge* deal, but the server-side bug really is impairing the usability of Gmail on Chrome right now.
Peter Kasting
Comment 23 2011-03-09 11:43:23 PST
(In reply to comment #22) > (In reply to comment #19) > > Peter et al, is there any chance that this fix is going to be pushed out soon? This isn't a *huge* deal, but the server-side bug really is impairing the usability of Gmail on Chrome right now. The fix has already been pushed for many people.
Jason Levine
Comment 24 2011-03-10 06:01:55 PST
(In reply to comment #23) > The fix has already been pushed for many people. On the off chance, is there something I need to do at my end to get my Google Mail session to "see" the new pushed stylesheet fixes? I've now logged into all four of my different Gmail accounts (personal, work, etc.), and am not seeing the fix on *any* of them.
Peter Kasting
Comment 25 2011-03-10 10:01:18 PST
(In reply to comment #24) > (In reply to comment #23) > > The fix has already been pushed for many people. > > On the off chance, is there something I need to do at my end to get my Google Mail session to "see" the new pushed stylesheet fixes? I've now logged into all four of my different Gmail accounts (personal, work, etc.), and am not seeing the fix on *any* of them. I'm sorry, I don't know the details of the gmail team's deployment plan. I would try clearing cache and cookies, and, failing that, fall back on using the stable build for a while.
Note You need to log in before you can comment on or make changes to this bug.