Subject: Re: [linux-audio-dev] an open letter to Linus re: low latency
From: Jay Ts (jay_AT_toltec.metran.cx)
Date: Fri Jun 23 2000 - 00:57:03 EEST
Stefan Westerfeld, stefan_AT_space.twc.de, Hamburg/Germany wrote:
>
> > example: why is it that enlightenment has to develop its own sound
> > server? kinda strange part of window manager...
>
> I think you can't blame anyone who is working on open source software (like
> Linus does) for not being interested in your-favourite-feature. People are
> developing the things they are good in and they are interested in.
Yes, very true. And yet the 'favorite feature' under discussion is something
that is vitally important to a broad class of programming applications, and
is a common application of desktop systems.
> So the *most important* thing which is required to have sound mixing support
> or low latency or whatever in the kernel is somebody who really cares about
> it, who likes working on it and who is good at it.
I've been paying attention to other things for a while, but if I understand
correctly, the limitation is that we aren't getting the low latency patches
included with the standard kernel, even as an option for building a new
kernel. It is Linus' role to accept or reject contributions from the
kernel developers. If he doesn't understand (or care) why a patch is
important, then he's likely to ignore it.
> I personally tend to think that having too much sound stuff in the kernel
> isn't a good idea either. We also do not run X11 in the kernel ;-).
Yes, I think we'd all agree. And yet Linux (and other Unix kernels) have
enough support for graphics to allow X to be possible. By copying Unix,
Linux got support for X "for free". However, sound cards are a relatively
new phenomenon, about as new as Linux itself. So Linux and other Unix
varieties are still in the process of accomodating the audio hardware in
a structured and sensible way. It took years for the Unix community to
agree on X as a standard, and I expect it will take a while to achieve the
same kind of goal for audio hardware.
And what I/we want to have included in the kernel is enough sopport for
audio that it's possible to write something like X11 for audio apps, and
have it work well enough to compare well with other platforms. There's
enough kernel support to make X possible for graphics, so why not support
sound, too?
Actually, as far as that goes, we have support in the Linux kernel for
graphics and sound, and right now, the real issue for both of them
is performance.
The folks at Precision Insight <http://www.precisioninsight.com> (with
help from the OS community) are working to add the Direct Rendering
Infrastructure to X and the kernel, to make 3D apps run fast, like they
do on Windoze. This is a work in progress, and things are happening.
IIRC, XFree 4.0 has a lot of that included with it. They are putting as
much as they can in user space, but there are still a few things that must
go into the kernel, to allow the X server to have a nearly direct access to
the hardware.
This is all very similar to DirectX in Windows, which supports both
graphics and audio.
So why not have the Linux kernel support high-performance audio as well?
If it requires kernel code to ensure low latency, then that will probably be
good for other purposes too (such as support for real time programming).
If audio applications (including ports of the power hitters like Cubase,
Cakewalk and Soundforge, etc.) are ever going to succeed on Linux, we
need to have things such that people can install them on an average
Linux system (e.g. the one they bought from Dell or IBM), without needing
a system administrator to fix the kernel for them. Most musicians are
not very good at doing things like that for themselves. Need I say more?
- Jay Ts
jay_AT_jayts.cx
This archive was generated by hypermail 2b28 : Fri Jun 23 2000 - 01:31:30 EEST