On Sun, 28 Nov 2004 13:50:40 -0500, Chris Metzler
<cmetzler(a)speakeasy.net> wrote:
On Sun, 28 Nov 2004 18:12:49 +0100
<SNIP>
Well, Matrox *does* suck. I bought a Matrox G550
because it was
an upper-end card with an open driver. I repeat: the open driver
was the biggest reason I bought a Matrox, and not another, card.
So what happened? I spent over a year trying and failing to get
a working 3D setup. What I got instead was constant hard lockups
requiring a full system reset, and X log files filling up with
error messages from the DRI drivers before the reset. Even with
journalling filesystems, that's unacceptable. Passing the info on
to Matrox merely got responses that OpenGL/3D under Linux was not
supported. Passing the info on to the DRI project folks got no
response at all; further queries revealed that the DRI project's
Matrox experts effectively aren't involved in the DRI project
anymore, so now no open source developers are working on the Matrox
drivers. A little bit of time spent in the
XF86/X.org/DRI bug
tracking systems will show similar Matrox bugs that don't get any
responses from the developers.
So what to do? I need halfway-decent 3D performance for my work.
Find a new line of work? Or buy a used old GF4 on ebay for $50?
<SNIP>
Until RMS can persuade people that the freedom to modify the software
one uses is as important as the freedom to work in the field of your
choice without being held back by race or gender or religion, people
and businesses are going to have a tough time justifying sacrificing
their financial security for that freedom.
It seems that RME has morphed into RMS. ;-)
Just because the driver is Open Source doesn't mean it works right. I
purchased an RME HDSP 9652 because it would be 'supported' under Alsa.
What I got was 80% support and a bunch of problems. Unfortunately the
developer, who worked amazingly hard to solve these problems was
unable to do so since he didn't actually have the card, and beyond
that, he didn't have the piece of external hardware I had that caused
the driver to fail.
Be careful what you ask for here. Information necessary to write a
driver is not the same as having a driver that really works. Unless
the developer has access to the hardware, and in the case of certain
sound card interfaces like spdif and ADAT, access to external hardware
that attaches to it, you may find yourself as I do. Owning a great
piece of hardware that only works 80%.
- Mark