DACP in Rhythmbox: Weekly Report 3



Hi all,

- What have I done this week?

This weekend has been very productive. During the week I had two
exams, and the time I took to think about my project didnt went very
well. I tried to think a way to get more logic into libdmapsharing
instead of Rhythmbox to discover new remotes, but in the end I could
only think about a solution that would add more code and not much
benefits.
But friday I was able to do much more. I changed the code from last
week to use G_TYPE_STRV instead of a GValueArray, which made the code
much cleaner.
I added a callback to remote pairing, so the UI will be able to detect
if pairing failed and show a message to the user. It was harder then I
thought, because user_data support in DMAPConnectionHandler wasnt
being used.
I fixed the db segfault cited last week and added signals for most of
DACP calls, still untested though.
Thanks to the Stereo program, I added all DACP structures, so we can
parse and send all DACP responses. Finally I added the correct
/ctrl-int response, so now iPhone recognizes it´s output and calls
more (unhandled) paths, which is really great and took a long time to
get right.

- What will I do next week?

I started documenting my work this week, so I should take a better
look at it this week. I should add more ctrl-int arguments this week,
finally testing if all the signals are correctly handled and implement
them in Rhythmbox.

- How accurate was my planning?

I did fix the DB segfault, was easier then I thought. I did
implemented basic /ctrl-int support, much harder then I thought and
the iPhone still doent show the player UI, which I thought it would.
But since it´s calling a bunch of unhandled paths, it´s clear why it
still doesnt.
I did a bunch of stuff which wasnt planned, but it was mostly fixes
and code cleaning.

- Cool stuff, problems?

I took a long time to understand code from Stereo, but it´s the only
open implementation of a DACP server. The Android Remote, as I found
out, doesnt call every DACP path, only the most useful ones, so it
wasnt very helpful this time.
It´s cool to compare Stereo Java implementation of DACP to
libdmapsharing GObject implementation. Both their advantadges and
shortcomings.

Until next week,
Alexandre Rosenfeld

Eng Comp 06 - USP São Carlos
FoG - http://fog.icmc.usp.br
IM Team - AIESEC


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]