[Bug 13422] Add nice API for channel creation

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=13422

--- Comment #14 from Guillaume Desmottes <guillaume.desmottes at collabora.co.uk> 2010-07-14 01:48:24 PDT ---
(In reply to comment #11)
> The request_ctx shouldn't need to have the TpDBusDaemon in it; I've added
> tp_base_client_get_dbus_daemon() in my trivia branch.

done.

> Shouldn't this async operation be cancellable? 

done.

(In reply to comment #13)
> If you don't convert request_ctx into a TpBaseClient subclass, its name should
> be in CamelCase anyway, because it's a type.

I'm not convinced having a subclass will buy us anything so I renamed it to
RequestCtx (I usually don't CamelCase such structure which are not an object).

> You seem to have a lot of non-breaking spaces in your docs, notably in
> tp_account_ensure_and_handle_channel_async. Please set your editor to highlight
> them - in particular, I suspect they break GObject-Introspection annotations.

TODO

> tp_channel_request_failed_cb() invalidates it with an error other than
> TP_DBUS_ERROR_OBJECT_REMOVED, and Succeeded with TP_DBUS_ERROR_OBJECT_REMOVED,
> so you don't need to watch for the individual signals at all - just use
> invalidated. The purpose of invalidated is to give you a single code path for
> all the possible reasons why your object caught fire and sank into the swamp
> :-)

fixed.

> Either tp_account_ensure_and_handle_channel_async or
> tp_account_ensure_and_handle_channel_finish or both should document that in
> particular, if the channel goes to a different handler, you'll see
> TP_ERROR_NOT_YOURS.

done.

> > tp_account_ensure_and_handle_channel_finish ()
> 
> Use foo() in gtk-doc, to make a hyperlink.

fixed.

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