[Bug 76120] [next] preparatory cleanup for GDBus migration

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=76120

Xavier Claessens <xclaesse at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |xclaesse at gmail.com

--- Comment #4 from Xavier Claessens <xclaesse at gmail.com> ---
 - "account-request test: don't leak account manager": I like adding in setup:

  g_object_add_weak_pointer (G_OBJECT (foo), &foo);

  and in teardown:

  g_object_unref (foo);
  g_assert(foo == NULL);

 - "tp_tests_proxy_run_until_dbus_queue_processed: don't use Introspect": I
don't like that hack, but I've got no better idea to workaround. Let's hope a
proper fix can go into GDBus.

 - "account-request test: prepare account manager": that really is a job for
the g_object_add_weak_pointer() trick I said above to be sure we are not
reusing the same singleton in next test.

 - "_tp_channel_got_properties: if we were invalidated, don't continue": Why do
we get the callback if the proxy is invalidated? I though that was one of the
difference between tp_cli_foo_call_bar() and g_dbus_connection_call(), the
former does not call callback and the later does call the callback but with a
GError set.

 - "tp_private_proxy_set_implementation: don't use g_assert_cmp* family": you
can use g_str_equal() when you know both strings are non-NULL.

 - "stream-tube test: don't assume one event per main loop iteration": In
create_tube_service() you should set test->tube_conns to NULL after
g_list_free_full() I think.

I tested here and make check pass, good job :-)

-- 
You are receiving this mail because:
You are the QA Contact for the bug.