[Webkit-unassigned] [Bug 142225] [GTK] WebView should hold a reference on WebContext because non-default contexts are a reality

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=142225

Michael Catanzaro <mcatanzaro at igalia.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mcatanzaro at igalia.com

--- Comment #4 from Michael Catanzaro <mcatanzaro at igalia.com> ---
Thank you!

It's not really clear from the test how it works, so I would add a comment at least. I guess the test crashes without your patch applied because the web view tries to use the context in dispose. So the test relies on an implementation detail, which is OK, but not great, especially since it is easy to test without relying on that. It also would be more obvious how the test works if you actually get the context from the web view after you've unreffed your copy of the context, and try to do something with the context.

Also not a bad idea to use assertObjectIsDeletedWhenTestFinishes() in a lifetime test, to test for a potential future leak. (Or an actual present tense leak, as Martin has pointed out. :)

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20150303/4c0622eb/attachment-0002.html>