Re: Thoughs about communication
- From: Matthew Hodgson <matthew matrix org>
- To: desktop-devel-list gnome org
- Subject: Re: Thoughs about communication
- Date: Fri, 27 Jan 2017 09:43:42 +0100
On 26/01/2017 22:52, Michael Catanzaro wrote:
It would be really awesome to have a GNOME Chat app based on Matrix.
Instead of implementing support for multiple protocols in the app,
like we did with Empathy, it would focus on doing one thing well --
Matrix, both text and video chat (OK, two things) -- and then the
quality of the support for other protocols would depend on the
quality of Matrix bridges and would not be something the app has to
worry about. Trying to support 20 different protocols really took its
toll on the Empathy user experience. Requires manpower. Maybe someone
will see this mail and become interested. Maybe not.
On the Matrix side: whilst I haven't played with it myself,
matrix-glib-sdk (https://github.com/gergelypolonkai/matrix-glib-sdk) was
looking like quite a good foundation for a GNOME Matrix client. As I
understand it, the only reason that Gergely stopped work on it was that
the impedance mismatch between Matrix & Telepathy was too great -
specifically, Telepathy lacked support for eventually consistent
infinite history which is one of the main defining properties of Matrix.
I suspect that resuming work there (potentially by Gergely, if he saw
interest from the wider GNOME community) is very much an option, and
obviously we'd support this as much as we can from the Matrix core team.
The Matrix client-server API has not changed much over the last 11
months, other than the addition of end-to-end encryption (although this
is still somewhat in flux whilst in beta).
Nachat (https://github.com/Ralith/nachat) and Quaternion
(https://github.com/fxrh/quaternion) show the potential for writing
native Matrix clients with Qt - and it'd be great to see a GTK equivalent.
Meanwhile, we'd be well up for running an official gimpnet Matrix<->IRC
bridge on matrix.org; we'll get one set up next week assuming that
sounds good. I'm assuming that if we connect via IPv6 we don't even
need a dedicated I:line.
thanks,
Matthew
--
Matthew Hodgson
Matrix.org
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]