Closed Bug 767822 Opened 12 years ago Closed 12 years ago

Failure in testGreyLarry.js | Timeout waiting for page loaded

Categories

(Mozilla QA Graveyard :: Mozmill Tests, defect)

defect
Not set
normal

Tracking

(firefox13 fixed, firefox14 fixed, firefox15 fixed, firefox16 fixed, firefox-esr10 fixed)

RESOLVED FIXED
Tracking Status
firefox13 --- fixed
firefox14 --- fixed
firefox15 --- fixed
firefox16 --- fixed
firefox-esr10 --- fixed

People

(Reporter: u279076, Assigned: remus.pop)

References

()

Details

(Whiteboard: [mozmill-test-failure])

Attachments

(3 files, 2 obsolete files)

Failure discovered during Firefox 14.0b9 functional tests.

Module: /testSecurity/testGreyLarry.js 	
Test: testGreyLarry.js::testLarryGrey 	
Error: controller.waitForPageLoad(): Timeout waiting for page loaded
Report: http://mozmill-ci.blargon7.com/#/functional/report/44fc451c2e6b66b62172f2e13e16a438
Attached patch patch v1 (default, aurora, beta) (obsolete) — Splinter Review
Now we use a local test page.
Assignee: nobody → remus.pop
Status: NEW → ASSIGNED
Attachment #636632 - Flags: review?(hskupin)
Comment on attachment 636632 [details] [diff] [review]
patch v1 (default, aurora, beta)

> /**
>  * Litmus Test 8806: Grey Larry
>  */

Can you please remove that part? We should really get rid of it at the beginning of the file.

>+  var domain = LOCAL_TEST_FOLDER.split("/")[2];
>+  controller.assertValue(webIDDomainLabel, domain);
[..]
>   controller.assertValue(webIDOwnerLabel, securityOwner);

Can you please change those asserts into expect.* calls? For the former I would propose expect.match() with a regular expression which check for the domain and port at the beginning of the URI.
Attachment #636632 - Flags: review?(hskupin) → review-
I don't see how .match will help us more then equal. We only must check if the accessed domain is the same as in the grey larry info page.
With match you wouldn't have to call 'split("/")[2];' and more traceable output would exist in the assertion message.
Addressed all requests.
Attachment #636632 - Attachment is obsolete: true
Attachment #639627 - Flags: review?(hskupin)
Attachment #639627 - Flags: review?(dave.hunt)
Comment on attachment 639627 [details] [diff] [review]
patch v2 (default) [checked-in]

Review of attachment 639627 [details] [diff] [review]:
-----------------------------------------------------------------

Looks good. Landed as:
http://hg.mozilla.org/qa/mozmill-tests/rev/d8c84efcec22 (default)
Attachment #639627 - Flags: review?(hskupin)
Attachment #639627 - Flags: review?(dave.hunt)
Attachment #639627 - Flags: review+
Status: ASSIGNED → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
Comment on attachment 639627 [details] [diff] [review]
patch v2 (default) [checked-in]

>-  controller.assertValue(webIDDomainLabel, "www.mozilla.org");
>+  expect.match(webIDDomainLabel.getNode().value, /\/\/(.*[^\/])/.exec(LOCAL_TEST_FOLDER)[1],
>+               "The domain label should equal the domain");

Just thought about it and we should better have used 'window.location.host' here. No strange regex would have to be used in this case.
Strange enough, controller.window.location.host returns "browser".
You probably want 'controller.window.content.location.host' or try to get it via the content document.
Attached patch patch v3 (default, aurora, beta) (obsolete) — Splinter Review
Henrik thank you for the tip. The problem is I can only send the value from controller.window.content.location.host if using a global variable. Otherwise that is not defined in the window that opens.
Attachment #640598 - Flags: review?(hskupin)
Attachment #640598 - Flags: review?(dave.hunt)
Comment on attachment 640598 [details] [diff] [review]
patch v3 (default, aurora, beta)

I'd like to see what Henrik thinks of this change. I'm not keen on storing the host as a variable, but other than leaving the regex I'm not sure what else to propose.
Attachment #640598 - Flags: review?(dave.hunt)
Comment on attachment 640598 [details] [diff] [review]
patch v3 (default, aurora, beta)

Hm, haven't thought about that this check is in a callback. So keeping the regex would make sense. Lets discard my last idea and get former patch backported to other affected branches. Remus, please check if it applies cleanly.
Attachment #640598 - Attachment is obsolete: true
Attachment #640598 - Flags: review?(hskupin)
Attachment #639627 - Attachment description: patch v2 (default, aurora, beta) → patch v2 (default, aurora, beta) [checked-in]
Applies cleanly on aurora and beta. Release and esr10 have to be fixed in regards to the favicon.
Attachment #639627 - Attachment description: patch v2 (default, aurora, beta) [checked-in] → patch v2 (default) [checked-in]
This needed the fix for favicon.
Attachment #641428 - Flags: review?(dave.hunt)
Attachment #641430 - Flags: review?(dave.hunt)
Comment on attachment 641428 [details] [diff] [review]
patch v2 (release) [checked-in]

Check your commit messages have the correct reviewer details.

Landed as:
http://hg.mozilla.org/qa/mozmill-tests/rev/cfea1b7916a9 (release)
Attachment #641428 - Flags: review?(dave.hunt) → review+
Comment on attachment 641430 [details] [diff] [review]
patch v2 (esr10) [checked-in]

Landed as:
http://hg.mozilla.org/qa/mozmill-tests/rev/27997d33fec6 (esr10)
Attachment #641430 - Attachment description: patch v2 (esr10) → patch v2 (esr10) [checked-in]
Attachment #641430 - Flags: review?(dave.hunt) → review+
Attachment #641428 - Attachment description: patch v2 (release) → patch v2 (release) [checked-in]
Product: Mozilla QA → Mozilla QA Graveyard
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: