Re: [Dri-users] Any Ideas How to Fix This

[Frank V. D. <fra...@st...>]

On Sunday 10 November 2002 21:02, magenta wrote:
> Could you point to any references about this?  I was under the impressi=
on
> that S3 was (back before they got bought out by whoever owns them now..=
=2E
> Diamond?) pushing it as an open standard.

All over the web, here is a short note among others:=20
http://dri.sourceforge.net/doc/faq/hardware.html .

> > > Now, if the original poster were willing to try the proprietary
> > > binary-only ATI drivers, those are much more likely to support UT2k=
3,
> > > but I have no idea what level of quality they are.  (I did see them
> > > running at Siggraph, but the booth jockeys only let it run a single
> > > demo which may as well just been a looping animation.)  They're
> > > available at
> > > http://mirror.ati.com/support/driver.html (I'd link to them directl=
y
> > > but that'd probably violate the license agreement, which is helpful=
ly
> > > put in a very narrow TEXTAREA with horizontal wrapping disabled).
> > Don't read it. Don't use it. Don't even go near.
>
> What, the license agreement?  You make it sound like it'll cause toads =
to
> rain down from the heavens and blood to rise from the sewers.

Well, there is just one point I made up for myself: I won't buy any more=20
hardware if there are no specifications open; I won't use any functionali=
ty=20
if it couldn't have been implemented by free indepentent software=20
devellopers. I also won't even be tempted to load a game that needs a cer=
tain=20
feature like s3tc, and no one will be able to seduce me to use "occasiona=
lly"=20
a closed driver "because it's just a small little thing".=20

I don't need companies like nVidia intruding in a free software world. Wh=
at=20
they do is embracing the whole linux/hardware acceleration thing.Do you w=
ant=20
Ati to go the same way? I did buy a radeon because I heard their specs ar=
e=20
half-open; they cannot open certain things they do not own theirselves bu=
t at=20
least someone will be able to get the specs, be it under an nda - it's th=
e=20
best you can get these days. I don't want to be the play-ball of yet anot=
her=20
company who thinks it can decide what to accelerate on my computer. dri i=
s=20
there so that the game doesn't have to matter about the underlying hardwa=
re,=20
right? These days, you get a logo smacked upon your face "NVIDIA-the way =
it=20
is meant to be played." Or something.=20

I once played ut2003 on a friends computer (under windows) and I admire s=
ome=20
aspects of the realism - but I won't run it. Not until they publish a pat=
ch=20
that removes the need for s3tc. Which might never be. I like to play a ga=
me=20
every now and then. Quake, urban terror, things like that. I would not li=
ke=20
to give that up. However if I find myself one day to buy certain hardware=
,=20
using patented technologies (legally or not!), running drivers or addon=20
software that is closed-source because some stinking company want to coun=
t=20
the tech in it to be hidden, or wants to know which people use that softw=
are=20
- I simply won't play. If this means the start of the game world being=20
devided between ati and nVidia games, fine. That won't be my world anymor=
e.=20
I'll use my expensive radeon card which plays quake at 90 fps for=20
dual-heading and tuxracer.=20

Call me stubbourn if you like. I see users, people on this list, thinking=
 by=20
themselves "ok, now we're into linux and free software and this kinda thi=
ngs,=20
with a bit of luck we're using a Matrox card, or an Ati (up to r200 - afa=
ik=20
Ati hasn't released anything about r250 or r300) so we even have opensour=
ce=20
drivers - life is good! Well for this ut2003 and s3tc thing... nah, I gue=
ss=20
it won't mean a disaster if /just once/ I use ati's own drivers. After al=
l=20
all windows users use Ati's drivers". Well at that moment you're locked i=
n to=20
a corporate principle (if that's good english) that says "come to me. All=
 you=20
game are belong to us."=20

About a year and a half ago, I published a text on my website explaining =
the=20
state of the nVidia drivers (which were even more unstable as today), and=
=20
explaining the dangers of closed hardware - you cannot make a driver if y=
ou=20
haven't got specs and there is NO WAY to do anything about bugs. You don'=
t=20
own your machine - your hardware manufacturer does. I also called to boyc=
ot=20
nVidia, and I called for solidarity and sympathy among windows users. Aft=
er=20
all, it's a pretty no-brainer for hardware firms to bother only about (or=
=20
focus on) windows users. So they are safe for good care from manufacturer=
s.=20
Yet, I did ask them for sympathy. By now, the page is a bit outdated,=20
especially concerning the actual quality of the driver - which is pretty=20
irrelevant to me. I don't care if any hardware firm even creates drivers =
for=20
ANY platform - as long as they don't deny others to do so.

Take nVidia. They spread their effort according to market share. Windows =
is=20
their primary goal, linux their second. They see a thing in linux and gam=
ing,=20
sure (money). Good thing? No, as you can read above. Why should they take=
=20
away the freedom from the user to use his hardware on a platform of his=20
choice? Suppose in 5 years the gaming market is windows-linux 50-50; and=20
nVidia, or Ati if they don't release specs of their newer models. Would y=
ou=20
feel good using their hardware? Yet now I think about users of more obscu=
re=20
platforms (some seem to still use beos, maybe one day free-or netbsd gets=
=20
something like dri) who will never be able to code a driver for their gef=
orce=20
- and at the moment, not for their radeon 8700 and better (and even then,=
 I=20
am asking myself if a beos user would be succesful asking for specs of th=
e=20
card I currently own).=20

Closed drivers? Just say no.=20

Frank

P.S. blessings to the DRI team.