From alsa-devel-owner@alsa.jcu.cz  Fri May 15 11:48:17 1998
Received: from pat.bath.ac.uk (qmailr@pat.bath.ac.uk [138.38.32.2])
	by marvin.jcu.cz (8.8.8/8.8.8) with SMTP id LAA28117
	for <alsa-devel@jcu.cz>; Fri, 15 May 1998 11:48:15 +0200
Received: (qmail 27786 invoked from network); 15 May 1998 09:48:11 -0000
Received: from solomon.bath.ac.uk (HELO bath.ac.uk) (mmdf@138.38.32.1)
  by pat.bath.ac.uk with SMTP; 15 May 1998 09:48:11 -0000
Received: from tobi.bath.ac.uk by solomon.bath.ac.uk id aa08818;
          15 May 98 10:48 BST
Message-ID: <355C0F43.7DE1@bath.ac.uk>
Date: Fri, 15 May 1998 10:47:47 +0100
From: "P.J.Leonard" <P.J.Leonard@bath.ac.uk>
MIME-Version: 1.0
To: alsa-devel@jcu.cz
Subject: Re: misc alsa things
References: <3.0.3.32.19980514163952.007cc530@mail.vt.edu> <3.0.3.32.19980514185849.00803ec0@mail.vt.edu>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: alsa-devel@alsa.jcu.cz
Sender: alsa-devel-owner@alsa.jcu.cz
Precedence: list

Alan Robinson wrote:
> 
> >On Thu, 14 May 1998, Alan Robinson wrote:
> >
> >> Userspace makes more sense to me, too, from the overall perspective, but
> >> might it not then be more lagged?
> >
> >Lagged in what sense? I think it will do fine for midi/mod playback ;-)
> 
> I am thinking in regard to realtime playback of a midi stream, such as from
> a MIDI keyboard, etc.


 Correct. If you try midi key board thru to synth in USER land you get
painful
timing problems when the system decides that it would like to
page out an X11 pixmap (or whatever). The midi thru must be
done within the kernel. This implies quite a lot of kernel code.
It should also be as modular as possible. For example a kernel
module might act as a mapper that connects a set of synth modules
to a set set of input modules. The mapper module would then need to
provide an interface to user land to allow the interogation of devices
and allow the routing of events. 

 My existing code is structured along these lines but it is all
implemented in user land and sits on top of the OSS stuff.
I will port the guts of my stuff to a kernel
module once the lower level device interface stuff is in place and the
kernel user land communication stuff is decided.


 cheers Paul.

