doc/TODO-matthias.txt
---------------------

IDEAS
-----

- publish TConversionTool in some way. Binary? Packaged with example programs?
- use speex (open source speech compression, http://speex.sourceforge.net/)
- use FLAC (free lossless audio codec)?
- check to shadow META-INF/services/javax.sound.* by means of prepending to boot class path.
- remove midishare and mp3 stuff?
- use libiiwusynth?
- support ADPCM?
- use TSE3 (portable sequencer engine written in C++)
- support TSE3MDL file format?

TODO (1 - important, 3 - nice to have)
----

general
2 - use aspectj!!!
3 - add "formatted for 80 cols" header to all files
3 - check freetts/Tritonus interaction

lowlevel, native
2 - initialize all variables
3 - use libtritonuscommon in all native libs
3 - rename DEBUG to debug_flag for all native libs
3 - separate static inner classes in all native/lowlevel classes
3 - in src/lib/alsa, change constants_check.h to constants_check.c and include it into the object list of in Makefile.in. This is to isolate its effects.
3 - change library building to using libtool
3 - re-check ALSA seq constants (constants_check.h)

documentation, distibution
1 - some documentation "how to use Tritonus", explaining general options (precompiled, compile yourself from cvs)
1 - make a note in INSTALL that for hand-installation only desired components should be installed. For automatic install, anyway. (see midishare exception)
1 - binary dists: esd, alsa, SRC, AOS
3 - cleanup web pages: contemporary doctype, weblint, ...
3 - Separate into tritonus_midi_alsa.jar and tritonus_sampled_alsa.jar
3 - review cdda distribution README

misc
1 - integrate ogg/vorbis
1 - test suite for classes (not necessarily interfaces) in javax.sound
1 - implement available() for lines
1 - fix start()/stop()/active behaviour for lines
2 - fix bug in AlsaMixer.readControlList()
2 - rework sample rate converter
3 - change provider search order to be exactely compatible with Sun



TEST
----

javax.sound.midi

Interfaces

ControllerEventListener
MetaEventListener
MidiChannel
MidiDevice
Receiver
Sequencer
Soundbank
Synthesizer
Transmitter

Classes

Instrument
MetaMessage
MidiDevice.Info
MidiEvent
MidiFileFormat
MidiMessage
MidiSystem
Patch
Sequence
Sequencer.SyncMode
ShortMessage
SoundbankResource
SysexMessage
Track
VoiceStatus
Exceptions 
InvalidMidiDataException
MidiUnavailableException


javax.sound.midi.spi

Classes

MidiDeviceProvider
MidiFileReader
MidiFileWriter
SoundbankReader


javax.sound.sampled

Interfaces

Clip
DataLine
Line
LineListener
Mixer
Port
SourceDataLine
TargetDataLine

Classes

AudioFileFormat
AudioFileFormat.Type
AudioFormat
AudioFormat.Encoding
AudioInputStream
AudioPermission
AudioSystem
BooleanControl
BooleanControl.Type
CompoundControl
CompoundControl.Type

Control
tested in ControlTestCase

Control.Type
tested in ControlTypeTestCase

DataLine.Info

EnumControl

EnumControl.Type

FloatControl

FloatControl.Type

Line.Info
LineEvent
LineEvent.Type
Mixer.Info

Port.Info

ReverbType

Exceptions 

LineUnavailableException
UnsupportedAudioFileException


javax.sound.sampled.spi

Classes

AudioFileReader
AudioFileWriter
FormatConversionProvider
MixerProvider

-------------------------------------------------------------------

SAOL
----

grammar:
- tables (s/r conflict)
- floats
- associativity of '?'
- make unicode-clean (but check if saol allows unicode chars in identifiers)


Tests mit JSInfo, Converter Test
================================
Es wird immer über die Funktion AudioSystem.isConversionPossible(AudioFormat, AudioFormat) getestet.

--------
source format: PCM_SIGNED, 1 channel(s), 16 bit samples, 2 byte frames, 44100.0 Hz, 44100.0 frames/second, little endian
target format: ALAW, 1 channel(s), 8 bit samples, 1 byte frames, 44100.0 Hz, 44100.0 frames/second, little endian
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.MpegFormatConversionProvider@ed1976dd
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.UlawFormatConversionProvider@f91976dd
possible: true
-
possible solutions:
- cancel alaw <-> ulaw completely (let smart converter do it)
- seperate it into an own converter
- derive Alaw-/UlawFCP from TMatrixFormatConversionProvider
---
source format: PCM_SIGNED, 1 channel(s), 8 bit samples, 1 byte frames, 48000.0 Hz, 48000.0 frames/second, little endian
target format: PCM_SIGNED, 1 channel(s), 24 bit samples, 3 byte frames, 48000.0 Hz, 48000.0 frames/second, little endian
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.MpegFormatConversionProvider@ed1a7a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.UlawFormatConversionProvider@f91a7a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.AlawFormatConversionProvider@fdd27a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.PCM2PCMConversionProvider@fcae7a60
possible: false
-
solution:
In PCM2PCM FCP., add 24 and 32 bit formats to set of I/O formats passed to super class constructor
additional suggestions:
a) remove duplicate lines in OUTPUT_FORMATS
b) rename OUTPUT_FORMATS to INOUT_FORMATS

---
source format: PCM_SIGNED, 1 channel(s), 8 bit samples, 1 byte frames, 44100.0 Hz, 44100.0 frames/second, little endian
target format: PCM_SIGNED, 2 channel(s), 8 bit samples, 2 byte frames, 44100.0 Hz, 44100.0 frames/second, little endian
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.MpegFormatConversionProvider@ed1a7a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.UlawFormatConversionProvider@f91a7a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.AlawFormatConversionProvider@fdd27a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.PCM2PCMConversionProvider@fcae7a60
possible: false
-
solution:
?? (this is presumably a fundamental solution of TEncodingProvider)
---
source format: PCM_SIGNED, 2 channel(s), 16 bit samples, 4 byte frames, 44100.0 Hz, 44100.0 frames/second, little endian
target format: PCM_UNSIGNED, 2 channel(s), 16 bit samples, 4 byte frames, 44100.0 Hz, 44100.0 frames/second, little endian
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.MpegFormatConversionProvider@ed1a7a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.UlawFormatConversionProvider@f91a7a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.AlawFormatConversionProvider@fdd27a60
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.PCM2PCMConversionProvider@fcae7a60
possible: false
-
solution:
discuss whether 16 bit unsigned is important
---
source format: PCM_SIGNED, 2 channel(s), 16 bit samples, 4 byte frames, 48000.0 Hz, 48000.0 frames/second, little endian
target format: PCM_SIGNED, 2 channel(s), 16 bit samples, 4 byte frames, 44100.0 Hz, 44100.0 frames/second, little endian
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.MpegFormatConversionProvider@d06e6b3f
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.UlawFormatConversionProvider@c46e6b3f
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.AlawFormatConversionProvider@c1566b3f
AudioSystem.doFormatConversionProviderIteration(): handling FormatConversionProvider: org.tritonus.sampled.convert.PCM2PCMConversionProvider@c0226b3f
possible: false
-
solution:
mal hinmachen!
---