[MPlayer-users] Re: best graphical card for mplayer

[D Richard Felker III]

On Wed, Sep 10, 2003 at 11:17:34PM -0500, Jonathan Rogers wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> D Richard Felker III wrote:
> >If they're playing games they should probably have a wintendo... I
> >don't see any point in putting a fancy 3d accelorator into a unix
> >system unless you're doing 3d modelling or something.
> >
> 
> It may come as a surprise to you, but there are actually quite a few 
> *nix users that play 3D accelerated games. The only reason I still have 
> Winders installed is for a couple of games, but I'd love to get rid of 
> it. Using Wine (and OpenGL), I almost can.
> 
> There's even a Nethack interface that uses OpenGL; I know it's 
> frivolous, but it's pretty cool. Do you think Nvidia and ATI develop 
> their 3D drivers primarily for modeling applications? I imagine there 
> are many ancient *nix admins that consider it silly to play movies on 
> "real" systems. Remember that Free Software is about freedom.

I was just basing what I said on two things:

1) There are essentially no free games that make good use of 3d
   acceleration. The few good free games are mostly text-based or 2d;
   all the 3d-accel-requiring ones I've seen have been very primitive
   and the models look like "my first 3dstudio project" talent level.

2) The non-free 3d accel games are mostly windows-only (altho some run
   under wine or have native ports). If you're going to put up with
   some buggy, crashing, non-free software (the games themselves), you
   might as well also be using windows. And that way they won't bring
   down your computer that you do real work on, just your games-only
   "wintendo".

Of course there are also previously-non-free games -- DOOM, Quake,
Duke3D, Descent, etc. which are "3d", but these run perfectly fine
(and much better imo) with software rasterization rather than 3d
accel, so I count them in with the "2d" games for purposes of
discussion.

> >Providing a driver is NOT a beneficial act on nvidia's part. Rather,
> >it reduces the demand for a free driver (because lots of fools will be
> >happy with the buggy x86-only proprietary one), discouraging people
> >from spending their time reverse engineering the hardware and writing
> >a driver because there's already one that's "good enough".
> >
> 
> I've always disliked using a proprietary driver, but is it even feasible 
> to reverse engineer hardware for functions as complex as 3D 
> acceleration? Obviously, Nvidia should release enough specs to 
> facilitate driver development. I wonder what it would take to bring this 
> about.

IMO legal action would be good. What other type of business, except
computer hardware, can get away with selling a product that requires
very specific technical information to use it, and NOT PROVIDING THAT
INFORMATION TO CUSTOMERS?!? Perhaps some users could try driving the
hardware themselves (making random guesses at how to control it) then
sue nvidia when they break it, since they weren't given sufficient
info to know how to avoid breaking it... :) Other ideas are welcome!


Rich