[Webkit-unassigned] [Bug 35942] [GTK] Fails http/tests/plugins/third-party-cookie-accept-policy.html

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Mar 10 05:40:54 PST 2010


https://bugs.webkit.org/show_bug.cgi?id=35942





--- Comment #7 from Dan Winship <danw at gnome.org>  2010-03-10 05:40:54 PST ---
(In reply to comment #4)
> (From update of attachment 50363 [details])
> What does SOUP_COOKIE_JAR_ACCEPT_NO_THIRD_PARTY mean?  Getting the proper
> behavior here (i.e., matching new-Safari) is somewhat subtle.
> 
> What cookie policy does Gtk want by default?  Is that cookie policy
> configurable by clients?

Sorry, didn't explain in detail because I figured Xan or Gustavo would be the
one reviewing it.

The issue here is that the cookie jar returned by defaultCookieJar() is really
only used by GtkLauncher and DumpRenderTree, because "real" apps (epiphany,
midori, etc) will call setDefaultCookieJar() to override it. So while epiphany
and midori were already providing a configurable no-third-party-cookie option,
the tests were still getting run with an accept-all-cookies policy.

(In reply to comment #6)
> About the patch, I'm not sure why it's stopping calling setDefaultCookieJar on
> the jar it creates, not sure that's right.

If you look at the old code, if you called defaultCookieJar(), and then called
setDefaultCookieJar() with a different cookie jar, the original default would
get leaked. The change makes it not leak.

(In reply to comment #3)
> Attachment 50363 [details] did not build on gtk:
> Build output: http://webkit-commit-queue.appspot.com/results/547002

> checking for LIBSOUP_2_29_90... no
> ...
> ../../WebCore/platform/network/soup/CookieJarSoup.cpp:41: error: ‘soup_cookie_jar_set_accept_policy’ was not declared in this scope

so, I guess the build would be fixed by adding some #ifdef HAVE_LIBSOUP_2_29_90
to the patch, but assuming that this is the build that would get used for the
tests, the test would still fail because it was using a pre-cookie-policy
version of libsoup. So I don't think we can unskip the test until the
build/test machines have a newer libsoup.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


More information about the webkit-unassigned mailing list