3D Support for NVIDIA

[Alan]

>
> Alan wrote:
> : Dean Messing wrote:
> : > Robin Laing wrote:
> : > : For Linux to get and maintain growth in the public and commercial
> eyes,
> : > : the developers have to work together to make these issues disappear.
>  It
> : > : would be great if all software was opensource but that isn't going
> to
> : > : happen in a world of get rich with IP plans in the works.
> : >
> : > Some people believe that they shd. get paid for the hard design work
> : > they do.  There's a great deal more (as I'm sure you know) to IP then
> : > "getting rich".  It's also about self protection from other rapacious
> : > companies, about portfolio negotiation, about protection from your
> : > ideas being stolen, and lots of other stuff.
> :
> : This brings up an interesting point...
> :
> : Last night at the PLUG Advanced Topics meeting in Portland Oregon one of
>
> Hi Alan. I've attended a few of those in years gone by.  Small world.

They we have probably met.  I am the one who organizes them and get speakers.

Want to talk at a PLUG Advanced Topics meeting?

> : the kernel developers who works at Intel talked about how they got the
> : lawyers to allow open sourcing of the ethernet drivers.
> :
> : The excuse they got from the lawyers was "intellectual property".  The
> : kernel developers said "what is it"?
> :
> : When the layers actually tracked that down, they found that there was no
> : IP in the drivers.  It was all in silicon.
>
> The issue is "layers of protection I think".  I, too, doubt that
> there's much real IP in the driver.

It was just that "IP protection" was the stock excuse without ever
stopping to thing about exactly what IP they were really talking about.

>
> : What the drivers do is communicate with an interface.  The rest should
> be
> : in hardware.  (Besides, anyone who really wanted to know would just
> attach
> : a bus analyzer and disassemble the drivers and figure out how it really
> : works.)
>
> It can be pretty hard if the driver has been run through a code scrambler.
> But, in principal, you are correct.

You don't expect that in a driver that is expected to be fast.  (Like a
video driver.)  Might explain Vista performance though... ]:>

> : > : I am going to
> : > : look at ATI in the future because of AMD's opening of their drivers.
> : >
> : > Ok, this answers my question above.  I didn't know that AMD had opened
> : > their drivers.  Good for them! Maybe that will put some good
> : > capitalist pressure on nVidia to do the same.
> :
> : Intel has opened their drivers as well.  Hopefully nVIDIA will do the
> : same.  I expect it will happen sometime shortly after the drivers are
> : completely reverse engineered. (So real soon now.)
> :
>
> I think as long as nVidia keeps innovating, their closed source driver
> will be a moving target.  What is interesting is that, at least before
> three years ago, one of the closed source driver authors, Mark
> Vojkovich, was also on the XFree86 Core Team.  So he made sure that
> the closed source driver worked well with X.
>
> I don't know what's become of him or if there's any connection btwn
> nVidia and XFree86 or Xorg these days.

The closed source people at nVIDIA have been pretty responsive.  The x.org
people have seemed a bit more pragmatic when it comes to dealing with
vendors.  (Whether that is good or bad is left to philosophical debate.)