[initial
hans@fugal.net**20050208010322] 
[skeleton
hans@fugal.net**20050208055036] 
[comment
hans@fugal.net**20050208055226] 
[jack_client_close is segfaulting
hans@fugal.net**20050208144748] 
[various
hans@fugal.net**20050208204946] 
[various
hans@fugal.net**20050209002503] 
[thread thoughts
hans@fugal.net**20050209051930] 
[more thread thoughts
hans@fugal.net**20050209052812] 
[event_sem
hans@fugal.net**20050210011230] 
[thread.h
hans@fugal.net**20050210062824] 
[iax_event_thread
hans@fugal.net**20050210063716
 Now I have jack and the iax_event helper thread producing, I just need to write
 the consumer. No doubt that will be the bulk of the work, but this is where the
 fun starts - I should start to see and hear exciting things soon.
] 
[try to make a call
hans@fugal.net**20050210151338] 
[thread work
hans@fugal.net**20050214200532] 
[capability!
hans@fugal.net**20050215005243
 Ah, the reason my side was rejecting the connection with "couldn't negotiate
 codec" was that mysterious capability argument to iax_call(). Apparently it's
 the codec capabilities that we can manage. That could be a bit clearer in the
 header file.
] 
[tests
hans@fugal.net**20050215005747] 
[top-level makefile
hans@fugal.net**20050215010635] 
[samplerate
hans@fugal.net**20050216003248
 code, but segfaults.
] 
[jack with libsamplerate
hans@fugal.net**20050216023036
 This should be working, although there isn't anything to listen to yet. I had a
 strange problem with alloca giving me a buffer that overlaps with the jack
 buffer. Not sure why. I don't really like mallocing a whole second worth of
 audio, but it should be sufficient for now.
] 
[We hear them
hans@fugal.net**20050216063122
 There's something wrong with the audio, but there is recognizable audio coming
 in and playing out the sound card. (Somewhere over the rainbow intro) Still not
 sending audio, that's the next step.
] 
[sending (awful) voice
hans@fugal.net**20050218003929
 Well, whatever is plaguing incoming voice is also plaguing us when we send
 voice. May be time to try a different codec (something like µlaw) and a
 different audio interface (e.g. alsa).
] 
[something very strange
hans@fugal.net**20050218053004
 When the following line in Jack::jack_process() is commented out, it works
 great (for incoming voice):
 jack_ringbuffer_write_advance(input_rb, (bp-buf)*sizeof(short));
 Otherwise it segfaults as soon as the input_rb ringbuffer fills up. output_rb
 get corrupted and it segfaults on src_short_to_float_array(). For the life of
 me I can't figure out what connection advancing input_rb would have on
 output_rb, but my best guess is that output_rb is adjacent to input_rb and
 something during advance input_rb is escaping.
 
 In happier news, I figured out why it sounded terrible. I needed to use the
 full API for libsamplerate and maintain state.
] 
[mysterious segfault solved
hans@fugal.net**20050218155235
 Thanks go to Magnus Hjorth for noticing what I did wrong.
] 
[design
hans@fugal.net**20050220190305] 
[various
hans@fugal.net**20050221182824] 
[refactor
hans@fugal.net**20050221030333
 Thus begins a major refactoring. True to the write one to throw away principle
 I have redesigned and am refactoring the code I can salvage in.
] 
[refactor iax
hans@fugal.net**20050221182054] 
[refactor jack and audio
hans@fugal.net**20050221182335] 
[refactor design
hans@fugal.net**20050221182703] 
[codec.h and protocol.h
hans@fugal.net**20050221184305] 
[resolved conflicts
hans@fugal.net**20050221184236] 
[redesign
hans@fugal.net**20050222063107] 
[refactor complete
hans@fugal.net**20050223052808
 The refactoring is complete. Funny, it doesn't look so much different than it
 did to start with. ;-)
 
 It compiles cleanly, but I haven't tested it yet.
] 
[design
hans@fugal.net**20050223053523] 
[various
hans@fugal.net**20050307231019] 
[nil
hans@fugal.net**20050309055730] 
[debugging still
hans@fugal.net**20050308063015] 
[progress
hans@fugal.net**20050309055307
 I think I'm making progress. I think I might be up against a brick wall on this
 laptop which seems to hate jack. It is finally communicating with the other
 end, according to ethereal, so now I can suspect the laptop as to why I don't
 hear anything. Either I need to implement ALSA or develop on falcon.
] 
[nil
hans@fugal.net**20050309055926] 
["goodbye"
hans@fugal.net**20050317052351
 We hear the lady say "goodbye" which is the asterisk equivalent of "Hello,
 world!" Still no voice transmitted. Also odd is not noticing the hangup.
] 
[take out the debug stuff
hans@fugal.net**20050307233824] 
[SIGINT
hans@fugal.net**20050317233404] 
[semrfd
hans@fugal.net**20050318000450] 
[nil
hans@fugal.net**20050318000520] 
[design refactor
hans@fugal.net**20050412042850] 
[refactoring
hans@fugal.net**20050412042911]