[linux-audio-dev] Re: Behringer

[Marek Peteraj]

On Thu, 2004-12-02 at 01:52, Jan Depner wrote:
> On Wed, 2004-12-01 at 20:15, Marek Peteraj wrote:
> > On Thu, 2004-12-02 at 00:29, Jan Depner wrote:
> > > On Wed, 2004-12-01 at 16:37, Dave Robillard wrote:
> > > > On Tue, 2004-30-11 at 17:43 -0500, Lee Revell wrote:
> > > > > No one said they were good.  I just said it was better than no support
> > > > > at all, and whatever RME decides to do, they designed the hardware, it's
> > > > > THEIR CHOICE.
> > > > 
> > > > No, it's not better than no support at all.  No support doesn't destroy
> > > > Linux in the long run.  Try to think on a little wider scale than
> > > > getting one silly little sound card to work in your specific (x86,
> > > > running a "supported" version of the Linux kernel) computer.  There are
> > > > more important things than trivial convenience for a small subset of
> > > > Linux users (at the expense of all the other ones) you know.
> > > > 
> > > 
> > >     My problem is a whole lot more important than 1 silly little sound
> > > card.  As I said before, somewhere around 200 Linux systems with NVIDIA
> > > cards and the proprietary driver.  The "more important" things you speak
> > > of are important to you but not to me.  I don't belong to your church.
> > 
> > Jan no offense. But i don't care about your 200 linux systems.  
> > Simply because if nvidia didn't care at all just like RME does with it's
> > fireface, you would use windows on your 200 machines. But OTOH, if you
> > had windows on those 200 machines before, had nvidia cards installed in
> > those boxes, and in order to reduce TCO you went with linux instead of
> > buying new licenses for a new version of windows, and were forced to use
> > nvidia binaries because those machines had nvidia cards installed
> > already, then that's kindof fine. Kindof because there still would need
> > to be a very _pragmatic_ reason to ditch old versions of windows. But i
> > certainly wouldn't advise people to go buy nvidia because of their
> > "exceptional" binary drivers. Too careless.
> > 
>     I certainly agree.  We're using windows because I write about 60% of
> the sonar/navigation processing software we use and I refuse to have
> Windoze in my office.

<snip>

>    I would like nothing better than to switch to an open source driver
> for exactly the reason you set forth.

If you need max performance then you have no choice. But for linux audio
it would be probably better to check out cards that are still available
and do have oss drivers and then recommend those. I believe that there
are some old radeon cards with passive coolers(isn't that what we, linux
audio users want ;) that do have oss drivers because ATI does provide
specs for their older cards(if they are still in production). At least
AFAIK, please correct me if i'm wrong. Also, perhaps if more and more
ATI customers went for their older cards, they'd certainly be forced to
rethink their policy. That said i asked them to release specs for their
latest GPUs, and i got a friendly 'no' response. The petition was at
~13000 signatures in Sept02. Today, 3 months later, it's ~20000 already.
I don't know how many ATI units are sold worldwide in one year but the
number must be huge compared to audio hw manufacturers. Still, 20000 is
an interesting number.   

I think it's the same case here. The one who will provide specs first is
going to secure his position in a new emerging market. 

Also, i can pretty much imagine the huge wave of new ultracool 3D demos
in the demoscene squeezing as much juice out of the GPUs as possible if
the registers & co documentation was released for a given GPU. 
So lots of positive things would happen. Let alone X.org and our new
accelerated desktops. Or perhaps samplerate conversion done on the GPU
in RT ;)  

That said, i can't buy older RME products anymore. The reason why i
freaked out on their forum is that after 15 products were supported in
alsa, after some people were cool enough to do the drivers which caused
them to sell say ~100 units(i think it's a lot more). What's more, they
provided support for it(helping people out, giving advices, fixing the
driver is what i call support). So they took all that off the shoulders
of RME. All for handing out a perhaps 15p pdf and 3-4 devices. All OK.
But. Tim Hall said, that we should build our relationship on trust
becasue it's a "two way process, which involves building trust". That's
exactly what triggered my responses(besides them being arrogant to other
customers in other threads).
" We don't share this information with anyone. " 
So after all this, they called these people "anyone". 
If i were the author of the driver i'd be very offended and i'd just pay
for the donated hw just to say.... ehm ;)

They had my trust. ;)

> > 
> > The point is not that someone might reverse engineer and do a
> > worse/better oss driver. The point is that nvidia, ati, xgi, matrox
> > *should* do open source drivers.
> > 
>     I agree but I can't make them even though we're a fairly large
> outfit.

I think that the oss community would be able to convince someone. But
then again, forget about politeness. It's a lesson i have learned.

Marek