From alsa-user-owner@alsa.jcu.cz  Tue Apr 28 12:08:28 1998
Received: from entry.jcu.cz (perex@entry.jcu.cz [160.217.1.111])
	by marvin.jcu.cz (8.8.8/8.8.8) with SMTP id LAA20941;
	Tue, 28 Apr 1998 11:53:25 +0200
Date: Tue, 28 Apr 1998 11:53:25 +0200 (MET DST)
From: Jaroslav Kysela <perex@jcu.cz>
To: alsa-devel@alsa.jcu.cz, alsa-user@alsa.jcu.cz
Subject: driver 0.1.0pre1 & utils 0.0.3
Message-ID: <Pine.LNX.3.96.980428111739.9476E-100000@entry.jcu.cz>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: alsa-user@alsa.jcu.cz
Sender: alsa-user-owner@alsa.jcu.cz
Precedence: list

Hi all!

	Looks that first stage of ALSA project is here. I fixed all known
bugs (by me - except persistent values support - kerneld/kmod - will be
in final 0.1.0). I hope that I solved all bugs for GUS MAX's codec
(please, confirm this) and for Sound Blaster 16/AWE. Please, report me now
all bugs. I'll release after few days version 0.1.0 of driver and work on
new sequencer should begin (some proposals are now out).
	In alsa-utils package is now very nice alsamixer from Tim Janik
with some small changes by me. This program shows how can be done
synchronization between multiple mixers or applications (try it!!!).
Package also contains updated amixer from Andy Lo A Foe. Big thanx to
both...
	New feature is support for raw (or direct) midi. I tried KeyKit
with my Roland XP-10 synthesizer (with GUS PnP) and it works again. SB
MIDI isn't tested yet.
	I have one big technical problem with MPU-401 and SB DSP. These
chips doesn't support transmit interrups, so output must be done in polled
mode (not good). If big data is present for output then system halts for
some time. I tried do some timing (fill FIFO - if possible - or transmit
only few bytes and suspend execution for some time). I'm using now system
timer (resolution HZ/100 = 10ms). It's not very good - looks like SB AWE
64 have only 2 byte FIFO (grrr). Compute - 100 * 2 = 200 bytes per second.
This is a little bit unacceptable. MIDI have transfer rate 31.250kbit/s.
31250 / 9bits per byte (include stop bit) = 3472 bytes per second (max).
Thus for SB AWE 64 is best timer resolution HZ/1736 = 576us :-(((( This is
unacceptable, too.. I'm thinking make some compromise (timing HZ/250 or
something this). User should change this timing value if it own some power
machine. Ideas? Looks like standards like SoundBlaster and MPU401 are very
bad for serious work. This is next reason why I like GUS cards.


						Jaroslav

Driver
======

0.0.12 -> 0.1.0

* next fix for stereo rate calculation for SB Pro (I hope last)
* fixes in GUS modules one DMA channel mode
* fixes in GUS Extreme driver (wrong control range)
* added alpha MIDI code for all supported soundcards except ES688 chip
  - GUS MIDI support
  - MPU-401 in UART mode support
  - SB MIDI support (not tested)
* Sound Blaster 16
  - fixed strange bugs in neutral fill which should cause system hangup
  - fixed real_rate computing
  - added /proc/sound/#/sb16 interface for DMA usage
    - 'echo "Playback 16" > /proc/sound/#/sb16' == use 16-bit DMA for playback
    - 'echo "Playback 8" > /proc/sound/#/sb16' == use 8-bit DMA for playback
    - 'echo "Playback auto" > /proc/sound/#/sb16' == auto DMA mode for playback
    - similar setup should be done for "Record" direction
* cleanups in event interface for mixer (for alsamixer which now works great)
* fixes in timing for CS4231 chip, GUS MAX should now work ok
* fixes in CS4231 module for one DMA channel mode

Alsa-utils
==========

0.0.2 -> 0.0.3

* included great ncurses alsamixer from Tim Janik <timj@gtk.org>
* enhanced Mixer from Andy Lo A Foe <arloafoe@cs.vu.nl> renamed to amixer

-----
Jaroslav Kysela <perex@jcu.cz>
Academic Computer Centre, University of South Bohemia
Branisovska 31, C. Budejovice, CZ-370 05 Czech Republic




