Richard Stallman wrote:
>I agree, it is high time that ALSA was made the standard sound driver
>in Linux. The existing sound module is limited; even worse, it
>advertises a proprietary freedom-subtracted software package as the
>remedy for its limitations.
>When such an important piece of free software as Linux contains ads
>for proprietary software, it gives the whole community an unfortunate
>message--not the sort of message will inspire people to write
>additional free drivers.
Well, I'm not so sure I agree with you here. (He said from UserSpace)
Quick story: I have a Soundblaster AWE64. I tried like the dickens to get it
working with the kernel sound drivers back around 2.1.98 or so, and nothing I
could do could get it fully functional. I could get parts of it working, but
not the whole thing.
Frustrated, I plunked down my cash for the commercial version of OSS - Ta daaa!
Functional sound, just like that. The only thing not supported is my joystick
port, and the fine folks at 4front tell me that's forthcoming.
Since then, I've more-or-less kept pace with the development kernels, and the
ppl at 4front have done the same, with a new version of the software I paid for
appearing within a day or so of each new dev kernel release. There have been a
couple of hiccups along the way (2.1.130-2.1.132 didn't work for me) but as of
2.2.0pre1, that's been fixed - with one noteable exception, Quake locks up
_hard_ when turning on sound. (Wasn't this fixed? Alan?)
While the move to avoid non-free software is certainly noble enough, those of
us in UserSpace have legacy software we want to keep working. Will ALSA provide
100% backwards compatability? Will it provide the same
ease-of-use/ease-of-configuration as the non-free software?
The current dev-tree-kernal Quake problems have been a royal pain in the ass.
In UserSpace, a kernel is just the overhead needed to make the apps run, and if
a kernel change breaks my app (and no fix for the app is available), then the
kernel is broken and unuseable. The only exception I'd make to this (somewhat
broad) statement, is when something in the kernel is so fundimentally wrong
that it's better to take the pain now then later (Glibc, a.out)
The switch to ALSA, IMHO, should only happen if and only if and when ALSA is a
near-transparent (to UserSpace) drop-in replacement for OSS.
Functionality first, nobility second.
DG
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/