GSOC: Android client potential
pidgin_m at fam-zangl.de
Sat Apr 13 17:20:07 EDT 2013
I did the porting last year and I am going to apply to do some more work
this year. I will be applying with two variants:
- One focusing on some more porting and writing the C/Java wrappers, so
that more of libpurple can be used from Java and from the Android UI.
This would also include basic plugin support, but not a mechanism to
install new ones.
- One focusing on the Android client, which means adding a real GUI
using fragments and adding a background service that will stay logged in
while the app is not displayed.
Since GSoC projects should not depend on each other and it would be a
bad use of resources to have two same projects, I would suggest for
anyone applying to consider that there may be multiple applications on
the subject. There are some areas that are mostly unrelated, that way we
could do multiple Android projects this year.
Relatively undepending projects include:
* Write a Chat client GUI that just works on a purple-like interface.
Libpurple does not need to be ported for this to work, but you could
only use the interface if you had libpurple or some sort of other
library that implements that interface. This interface could then be
used by other projects in the future. This work would only include Java
work, so it would be a good project for someone not familiar with C (but
if you want to start on this, you should be good at understanding
* Work on multimedia support for Android. Voice should be pretty simple,
but I have not yet looked into the video APIs of Android. This can be
tested with a simple test app and also does not need the other Chat
functionality. The result should be the adapted library that can then be
integrated into an Android GUI.
* Write something that relays messages from a normal PC to Android, to
not need Android staying connected to all the accounts. This could be
done by writing a stand-alone client for Android and a plugin for
pidgin. Or it can be done by adapting libpurple to support this sort of
synchronisation, so that you do not need an Android client for testing,
but it will work on Android as soon as libpurple works on Android. This
would then also be a nice project for somebody who is not familiar with
the JNI or Android API.
* Work on Plugin support for Android. Android plugins would need a
special sort of delivery/unpacking/including. Currently, all protocols
are just compiled statically. This could also include adding a new sort
of Java-Plugins that can then include bigger UI changes and Android API
access (you could also do this form C code, but using JNI for every
simple bit is not really nice). The aim could be to support some common
plugins, like Encryption. It would also be nice that all the protocols
work, currently only Jabber is usable (not stable ;-)).
If you are a student and want to participate in one of the above
projects, feel free to ask me questions.
I would really like having a stable and good Android client in a few years.
Am 13.04.2013 22:20, schrieb Peter Lawler:
> Just throwing this out into the ring.
> Last year, some GSOC work was done to bring the core of Pidgin -
> libpurple - to Android. It's reported as working, albeit with a very
> basic UI. If I recall correctly, it's currently at libpurple 2.10.6, one
> micro release behind current.
> I would encourage all students contemplating applying to Pidgin to
> consider working on Pidgin for Android (Pigdroid?)
> Whilst I can't speak for the team who will approve applicants, I'm
> merely a semi-retired retired Crazy Patch Writer, I do suspect and hope
> that a good plan to progress this would be a strong candidate for
> acceptance. There seems to be much interest amongst users and (afaik)
> some of the developers in getting this further along.
> More details on last year's GSOC Android effort can be found
> https://developer.pidgin.im/wiki/GSoC2012/Android - I encourage people
> to look at it.
> Devel mailing list
> Devel at pidgin.im
More information about the Devel