686129 - Random orange: TEST-UNEXPECTED-FAIL | test-1-introduction.js | test_open_and_close_migration_assistant and other related failures

Status: RESOLVED FIXED

(Thunderbird :: Testing Infrastructure, defect)

Description

[Rain [:rain1] (inactive)]

Attachment: patch v1

e.g. http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1315665729.1315666674.10002.gz

Attaching what I think is a fix for the issue. It's a departure from our usual way of wait-then-verify, but I think it's appropriate in this case. I've unified all the code that waits for either a content tab or a frame to load as far as I could.

Comment 1

[Andrew Sutherland [:asuth] (he/him)]

Comment on attachment 559698 [details] [diff] [review]
patch v1

The key thing is that we're reasonably confident the thing we expected to happen actually happened.  This way (one call instead of two) is definitely an improvement, especially since if people don't cargo cult correctly, they might never put in an assertion after the first thing!

http://reviews.visophyte.org/r/559698/

on file: mail/test/mozmill/shared-modules/test-migration-helpers.js line 86
>   if (!aPane)
>     return fc;

Should this be [fc, null] in keeping with the new return signature?

Comment 2

[Rain [:rain1] (inactive)]

Nice catch!

Comment 3

[Rain [:rain1] (inactive)]

Comment on attachment 559698 [details] [diff] [review]
patch v1

clearing review request because I think asuth's review is enough.

Comment 4

[Rain [:rain1] (inactive)]

https://hg.mozilla.org/comm-central/rev/c8adaba7bea5

Comment 5

[Rain [:rain1] (inactive)]

oops, forgot to check in the review comment. https://hg.mozilla.org/comm-central/rev/6538067b8784