On Thu, Jan 18, 2001 at 09:30:06AM +0100, Christoph Reichenbach wrote:
> BTW, we need a stable sound subsystem for the release. If you intend to
> break something that will take a long time to fix, then we should fork off
> a stable CVS branch.
I believe 0.3.1 should go out without a "working" sound system, Note
that I define "working" as actually playing music. :) And that will be
a hell of an accomplishment.
So let me explain myself a bit more: All of the work I plan to do in
the immediate future is on the sound framework -- In other words,
everything but the actual calls to mt32_play_note().
My immediate plans:
1) config.l & main.c recognise and set up all of the sound config
options
2) PCM and MIDI output api codified.
3) Since the game state doesn't depend on what the actual sound hardware
is, This configuration will not be part of snd_state or game_state,
and instead part of the soundserver. The only gamestate thing is the
sound volume, and that's already there. *note to self -- make sure
the savegame code saves the volume*
5) Write new null PCM and MIDI output drivers that do nothing but
optionally print out the commands the soundsever sends to it. An
excellent way of making sure the API is in place and otherwise functions.
5) Rename the soundserver_null to soundserver_unix. It will handle all of
the actual sequencing and event handling. Modify it to call the
midi_driver->playNote() or whatnot to actually make things happen.
Out of all of this, there really isn't anything that is worth holding
back 0.3.1 for, since actual sound output won't take place until the
rest of the stuff is in place. Only step 5 has the potential to
introduce breakage and instability in what a user will see... because
until I hit that point, all of my changes are outside the game loop.
Whee.
- Pizza
--
Solomon Peachy pizzaATfucktheusers.org
I ain't broke, but I'm badly bent. ICQ# 1318344
Patience comes to those who wait.
...It's not "Beanbag Love", it's a "Transanimate Relationship"...
