594687 - [SeaMonkey 2.1, mochitest-browser-chrome] TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/suite/common/tests/browser/browser_346337.js and browser_466937.js

Status: VERIFIED FIXED

(SeaMonkey :: Testing Infrastructure, defect)

Description

[Bruno 'Aqualon' Escherl]

TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/suite/common/tests/browser/browser_346337.js | The value for "//input[@type='file'][1]" was correctly restored
TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/suite/common/tests/browser/browser_346337.js | The value for "//input[@type='file'][2]" was correctly restored

Seems only to occur on Windows: http://brasstacks.mozilla.com/topfails/test/SeaMonkey?name=chrome://mochikit/content/browser/suite/common/tests/browser/browser_346337.js

Comment 1

[Bruno 'Aqualon' Escherl]

We're missing the test changes from http://hg.mozilla.org/mozilla-central/rev/8d846fde08cb (bug 583863)

Comment 2

[Bruno 'Aqualon' Escherl]

Attachment: update tests for refactored input implementation

This are the test changes from bug 583863. The test passes for Linux with and without the patch, so I couldn't test it myself.

Comment 3

[Bruno 'Aqualon' Escherl]

TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/suite/common/tests/browser/browser_466937.js | normal case: file path was correctly preserved - Got , expected /home/user/regular.file

The patch should also fix that test fail.

Comment 4

[[e-mail]]

Comment on attachment 473468 [details] [diff] [review]
update tests for refactored input implementation

Probably the paths previously used weren't valid on Windows, but it didn't matter because nobody validated them before.

Comment 5

[[e-mail]]

Comment on attachment 473468 [details] [diff] [review]
update tests for refactored input implementation

On second thoughts, r=me if you remove all the calls to createUnique.

Comment 6

[Bruno 'Aqualon' Escherl]

Ok, updating the patch (removing checkin-needed until I have an updated patch)

Comment 7

[Bruno 'Aqualon' Escherl]

Attachment: updated tests without the calls to createUnique

(In reply to comment #5)
> Comment on attachment 473468 [details] [diff] [review]
> update tests for refactored input implementation
> 
> On second thoughts, r=me if you remove all the calls to createUnique.
Taking over r+

Comment 8

[Serge Gautherie (:sgautherie)]

Comment on attachment 473485 [details] [diff] [review]
updated tests without the calls to createUnique

Could we use ProfD instead of TmpD, fwiw?

Comment 9

[Bruno 'Aqualon' Escherl]

Does it make any difference?

Comment 10

[Justin Wood (:Callek)]

For future:
http://blog.bonardo.net/2010/06/22/so-youre-about-to-use-checkin-needed

Which you do a decent job of, only comment is (especially if you are already doing a final upload) is specify bug # and specify who reviewed.

That said, pushed as: http://hg.mozilla.org/comm-central/rev/3747f1d60f0a

Comment 11

[Serge Gautherie (:sgautherie)]

(In reply to comment #9)
> Does it make any difference?

It helps to keep TmpD clean, and to put all test files in the same place.

*****

V.Fixed, per
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1284108915.1284114436.22079.gz
WINNT 5.2 comm-central-trunk debug test mochitest-other on 2010/09/10 01:55:15