Bug 42625

Summary:

sputnik S15.9.3.1_A5_T1~T6 fail when run in timezone other than Pacific Time

Product:

WebKit

Reporter:

Xianzhu Wang <phnixwxz>

Component:

Tools / Tests

Assignee:

Zan Dobersek <zan>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

abarth, ap, bweinstein, commit-queue, dpranke, eric, ojan, webkit.review.bot, zan

Priority:

Version:

528+ (Nightly build)

Hardware:

All

OS:

All

Attachments:

Description	Flags
Patch	none
Patch	eric: review-

Description Xianzhu Wang 2010-07-20 03:34:20 PDT

On a machine whose timezone is not Pacific Time, the following tests fail:

fast/js/sputnik/Conformance/15_Native_Objects/15.9_Date/15.9.3/S15.9.3.1_A5_T1.html -> failed
.
fast/js/sputnik/Conformance/15_Native_Objects/15.9_Date/15.9.3/S15.9.3.1_A5_T2.html -> failed
.
fast/js/sputnik/Conformance/15_Native_Objects/15.9_Date/15.9.3/S15.9.3.1_A5_T3.html -> failed
.
fast/js/sputnik/Conformance/15_Native_Objects/15.9_Date/15.9.3/S15.9.3.1_A5_T4.html -> failed
.
fast/js/sputnik/Conformance/15_Native_Objects/15.9_Date/15.9.3/S15.9.3.1_A5_T5.html -> failed
.
fast/js/sputnik/Conformance/15_Native_Objects/15.9_Date/15.9.3/S15.9.3.1_A5_T6.html -> failed

Comment 1 Alexey Proskuryakov 2010-07-20 10:13:02 PDT

We should probably just disable those tests then.

Comment 2 Eric Seidel (no email) 2010-08-17 10:21:09 PDT

Agreed.

Comment 3 Eric Seidel (no email) 2010-08-17 10:28:04 PDT

Created attachment 64605 [details]
Patch

Comment 4 Alexey Proskuryakov 2010-08-17 10:48:30 PDT

Comment on attachment 64605 [details]
Patch

r=me.

Comment 5 WebKit Commit Bot 2010-08-17 11:07:21 PDT

Comment on attachment 64605 [details]
Patch

Clearing flags on attachment: 64605

Committed r65519: <http://trac.webkit.org/changeset/65519>

Comment 6 WebKit Commit Bot 2010-08-17 11:07:25 PDT

All reviewed patches have been landed.  Closing bug.

Comment 7 Zan Dobersek 2012-08-02 23:00:07 PDT

Could these tests be skipped if the system's timezone is incorrect? Similarly to how some tests are skipped based on missing symbols in modules.

Comment 8 Zan Dobersek 2012-08-23 10:33:44 PDT

Reopening to attach new patch.

Comment 9 Zan Dobersek 2012-08-23 10:33:47 PDT

Created attachment 160193 [details]
Patch

Comment 10 Alexey Proskuryakov 2012-08-23 10:50:10 PDT

Honestly, I don't think that having these tests enabled on some machines is worth the complexity.

We should just remove sputnik tests and migrate to ecma-262 official suite. Maintaining that seems more worth the effort long term.

Comment 11 Zan Dobersek 2012-08-23 10:54:22 PDT

(In reply to comment #10)
> Honestly, I don't think that having these tests enabled on some machines is worth the complexity.
> 
> We should just remove sputnik tests and migrate to ecma-262 official suite. Maintaining that seems more worth the effort long term.

I'm supportive of the migration, but these tests (or their equivalents) are still present in the ecma-262 suite though.

Comment 12 Dirk Pranke 2012-08-23 10:55:34 PDT

Is there a way to file bugs or submit patches to the official suite to get these fixed?

Comment 13 Eric Seidel (no email) 2012-08-23 11:02:46 PDT

Comment on attachment 160193 [details]
Patch

We could also just delete them.  I suspect this will just end up being confusing. :)

Comment 14 Alexey Proskuryakov 2012-08-23 11:26:38 PDT

My suggestion is to keep these tests disabled unconditionally, and try to get official tests fixed.

Comment 15 Eric Seidel (no email) 2012-08-23 11:43:55 PDT

Comment on attachment 160193 [details]
Patch

I don't think it really matters what we do with these tests.  I think the status quo is fine, and this patch isn't really worth the complexity.  I appreciate you looking at the problem!

Comment 16 Zan Dobersek 2012-09-03 23:57:08 PDT

(In reply to comment #12)
> Is there a way to file bugs or submit patches to the official suite to get these fixed?

The bug's already reported in the Ecmascript Bugzilla:
https://bugs.ecmascript.org/show_bug.cgi?id=330