[linux-elitists] Laptop that doesn't suck?

Jason Spence jspence@lightconsulting.com
Wed Jan 21 09:51:33 PST 2004


On Tue, Jan 20, 2004 at 02:42:12PM -0800, Ben Woodard wrote: 
> On Mon, 2004-01-19 at 20:56, Don Marti wrote:
> > begin Greg KH quotation of Mon, Jan 19, 2004 at 02:05:20PM -0800:
> > 
> > > You are saying that NVidia has code that infringes on another company's
> > > IP in their closed source module?  Oh, that makes me feel even better...
> > 
> > I asked them about this, and the answer is basically that they know
> > there are a hell of a lot of graphics patents out there, hardware
> > and software, valid and invalid, and if they published driver source
> > it would make it easier for a patent holder to make a case, bogus
> > or not, against them.
> 
> >From what I've heard NVidia say that is being rather kind. Basically
> they know that are infringing in a lot of areas. They also know that
> several other companies are infringing on their patents and so the only
> thing that allows progress to move forward in that whole arena is a sort
> of realization of a state of mutually assured destruction. However, the
> fact that their opposing patent portfolios are the only think that
> protects them against each other also makes it such that is virtually
> impossible for another to enter the market space. 

Oh, so thaaaats what's going on.  The entire situation makes a whole
lot more sense now from that perspective.

> IMHO this situation almost proves that having large "defensive" patent
> portfolio idea works from a business point of view but demonstrates how
> this philosophy ultimately is anti-competitive.

The scenario I keep hearing repeated when I bring this up goes
something like:

1) I spend X on expensive engineers and R&D
2) They come up with neat ideas
3) I put them in software
4) The competitors bust out their disassemblers and steal my ideas
5) As a result, they can price their products lower than mine because
   they didn't have to spend X in addition to other product
   development costs, and I can't sue them because I don't have a patent.

> The other thing that was startling is that their engineer admitted that
> ATI has better hardware but they manage a slight edge due to their
> software.

ATI's programmers couldn't produce quality software if their life
depended on it.  I've had to deal with the software packages for
several of their products and they all have weird and annoying
quirks.  My favorite was how the DirectShow capture interface for
their latest All-In-Wonder card would lie about the format of the data
stream, causing all the applications to display the image stretched or
in funky colors.  The best part is that it would do so for formats
that were supposed to be tested by WHQL before they should have been
able to get their driver signed by MSFT!

Oh, and the fact that their OpenGL library is embarassingly unstable
in selection mode.  And that parts of it are apparently implemented in
the kernel, which is why I was suddenly so familiar with the
IRQL_NOT_LESS_OR_EQUAL bugchecks when I started testing some of my
stuff to their cards.

And don't even get me started on their Linux drivers; I once wrote an
x86 multiboot microkernel for the IOCCC on a drunken bet and I swear
it was more stable than that pile of crap.

-- 
 - Jason                     Currently at: Ohlone College (Fremont, CA) (Fair)

Taxes are going up so fast, the government is likely to price itself
out of the market.



More information about the linux-elitists mailing list