1066613 - dom/documents/resource-metadata-management/document-cookie.html webplatform test is going to perma-fail when Gecko 35 merges to beta

Status: VERIFIED FIXED

(Core :: Networking: Cookies, defect)

Description

[Ryan VanderMeulen [:RyanVM]]

The webplatform tests are due to ride the Gecko 35 train. Per the run below, the document-cookie.html test will perma-fail when it hits beta.

Jason, I'm not even sure who's working cookies at this point. Any suggestions for someone who can look into this?

https://tbpl.mozilla.org/php/getParsedLog.php?id=47848275&tree=Try

21:57:47     INFO - TEST-START | /html/dom/documents/resource-metadata-management/document-cookie.html
21:57:48     INFO - TEST-PASS | /html/dom/documents/resource-metadata-management/document-cookie.html | document has no cookie 
21:57:48  WARNING - TEST-UNEXPECTED-FAIL | /html/dom/documents/resource-metadata-management/document-cookie.html | getting cookie for a cookie-averse document returns empty string, setting does nothing - assert_equals: expected "" but got "test=foobar"
21:57:48     INFO - TEST-OK | /html/dom/documents/resource-metadata-management/document-cookie.html | took 263ms

Comment 1

[Jason Duell]

Cookies are under-owned right now, but I'm cc-ing the usual suspects.  (I just fixed bug 966713 and my physician told me I should only eat one cookie bug a day :)

Comment 2

[Ryan VanderMeulen [:RyanVM]]

Still permafailing. Any news?

James, I guess our other option is to just mark it as known-fail once this merges to beta?

Comment 3

[James Graham [:jgraham]]

Yeah, in general I don't have a better suggestion than changing the expected results once we merge across branches. It isn't easy to get the pref values out, so it isn't easy to make the test results depend on these values.

Comment 4

[Josh Matthews [:jdm]]

Sorry I didn't address this before. I also have no suggestion besides changing the expected results.

Comment 5

[Josh Matthews [:jdm]]

Actually, I'm confused - is this test not failing on mozilla-central, but will fail after the merge? Do we know why?

Comment 6

[Josh Matthews [:jdm]]

Oh, cookieBehaviour = 0 on non-mozilla-central. Yeah, all we can do is adjust the expected values, unfortunately.

Comment 7

[Ryan VanderMeulen [:RyanVM]]

Confirmed fixed by bug 1046430.