Hi all,
I'm glad to welcome the list members and hope that our joint work will be
interesting and useful for both Sun and open-source community.
Let me overview current status of JavaSound.
As you know currently whole JavaSound implementation is closed. Unfortunately it
was late decision (to close JavaSound) and there was no time to make it
"partially" opened.
So #1 goal for now is split JavaSound into 2 parts (closed & opened) and
open-source last one.
Open part will include the following:
- "direct audio" mixer implementation (DirectSound/ALSA/Solaris Audio Mixer);
- Ports mixer implementation (WinMM/ALSA/Solaris Audio Mixer);
- Midi i/o devices implementation (WinMM/ALSA);
- various file readers/writes, format converters.
Closed part will include all Beatnik code
Simple dropping this code will cause too much feature decline, so we have to
make alternative implementation of some features before dropping old one.
What we need:
1) Software synthesizer.
Current one support only .gm soundbanks (beatnik proprietary format) and does
not support "industry standards" like .sf2. This task is big enough (in addition
to Synthesyzer implementation it requires to implement set of related
interfaces/classes). This task is not related with currently closed JavaSound
classes so it's already opened as separate open-source project
(audio-engine-dev; unfortunately name is not informative enough);
2) OSS mixer (for linux/solaris)
I think OSS has a big chance to become Solaris standard audio engine so
JavaSound have to handle it (both sampled and midi).
After implementing this 2 features it will be possible to drop closed part at
all. JavaSound will drop support for .gm soundbanks and .rmf files (both formats
are close (not public) and are not widely used), "Java Sound Audio Engine" mixer
(HeadspaceMixer).
As soon as part of JavaSound implementation will be opened, it will be possible
to start working on other issues (bug fixes and enhancements).
Regards
Alex
