[Telepathy] Telepathy and EFL

[Kai Huuhko]

Sat, 11 Feb 2012 12:27:20 +0200
Kai Huuhko <kai.huuhko at gmail.com> kirjoitti:

> Fri, 20 Jan 2012 17:53:17 +0200
> Kai Huuhko <kai.huuhko at gmail.com> kirjoitti:
> 
> > Mon, 16 Jan 2012 13:53:11 -0500
> > Youness Alaoui <youness.alaoui at collabora.co.uk> kirjoitti:
> > 
> > > On 01/15/2012 06:51 PM, Danielle Madeley wrote:
> > > >> 20:38 <kuuko> Hello, I'm looking into rewriting a python app
> > > >> that's using the low level telepathy functionality. I
> > > >> understand that telepathy-python has been deprecated in favor
> > > >> of telepathy-glib python bindings. The thing is that the app
> > > >> is not using GTK+ for it's ui, instead it's EFL-based, so I am
> > > >> wondering if and how I can make use of telepathy-glib in such
> > > >> a client.
> > > > 
> > > > Your problem with using the telepathy-glib bindings will be
> > > > integrating with the GLib mainloop to receive events and
> > > > callbacks. I don't specifically know about EFL, but nowadays Qt
> > > > and GLib share a mainloop.
> > > > 
> > > > Possibly this blog post is useful to you:
> > > > http://blog.gustavobarbieri.com.br/2009/10/01/ecore-glib-main-loop-integration/
> > > > 
> > > > If you can't integrate the mainloops, you may have to stick with
> > > > the low-level telepathy-python bindings,
> > > > 
> > > 
> > > Yes, ecore integrates with the mainloop from glib so that
> > > shouldn't be a problem. It would be great to have an EFL based
> > > client for Telepathy. If you're interested, we (the aMSN team)
> > > already started (a few years ago) working on a python based
> > > client and you could easily port it to use telepathy instead of
> > > directly using papyon. I've wanted to do that for a while but
> > > haven't had time. We have an EFL front end for aMSN2 although I'm
> > > not sure of its current status. Interested ?
> > 
> > I shall take a look. Can you provide me with links to said clients?
> > 
> > For the project at hand, what I am interested in is specific
> > documentation and example code for telepathy-glib in python.
> > 
> > For now, I'm using as much of the existing code as possible,
> > changing the direct connection handling to MC5 specific features
> > such as account/connection handling through AM and channels through
> > CD, using telepathy-python client bindings.
> > 
> > It is only once I feel having sufficient fluency with the current
> > code base and tp-glib that I am prepared to begin the move to the
> > latter bindings.
> > 
> > The application I am working on is Apathy:
> > <http://www.openapathy.org/> I am not paid or sponsored by anyone
> > for this work.
> > 
> > The time table I have set for the initial work is until the end of
> > January. At that point I shall reassess the situation.
> > 
> > Thank you for your answers and interest.
> 
> I have begun the rewrite and solved the first obstacles that came up.
> However this has me stumped.
> 
> Fetching a contacts avatar with get_avatar_file(), I get:
> 
> TypeError: metaclass conflict: the metaclass of a derived class must
> be a (non-strict) subclass of the metaclasses of all its bases
> **
> ERROR:pygobject.c:929:pygobject_new_full: assertion failed: (tp !=
> NULL)
> 
> and the whole process dies.
> 
> I have in my factory Tp.ContactFeature.AVATAR_DATA, and get the token
> just fine. Am I missing something here? Must I listen for
> notify::avatar-file or do something else before trying to fetch that
> property? This happens with gabble; with haze I get None for all
> avatars.

I found what the problem was, it was my mistake: I had left in import
tpfarsight, farsight from legacy code.