Possibly offtopic : Binary only driver

[Mike Hearn]

On Sun, 21 Nov 2004 16:29:12 +0100, Arjan van de Ven wrote:
>> > At least that's the intention. Some people think they can get away with
>> > not being GPL while still using and depending on deep kernel internals;
>> > well... to be honest the pain is on them though.
>> 
>> Actually the pain is on the users. You just don't see it.
> 
> people who choose to use binary modules suffer, I am fully aware of
> that. I rather not have them suffer that. But in a way they choose that
> pain when they choose to start using binary modules.

Sometimes the choice is like this.

I bought a 3D game. I want to play it.

a) Go back to Windows
b) Use a binary driver on Linux

> But look at it with a slightly longer pervue; about half the bugs
> reported against the kernel are in the drivers. The fact that the driver
> sources are available gives us a shot at fixing those, just as the
> availability of the core source allows us to fix bugs there.

Well this is just a generalisation of "all software should be open
source". I tend to agree with that, it would be great if that were true.

But it's not true, and there are no signs of it suddenly becoming true
anytime soon. I'd also rather people open sourced their code due to the
social benefits (everybody being able to fix bugs, share knowledge,
implement new system-wide optimisations etc) rather than because the
kernel made it a total pain in the ass to do otherwise. That's coercion
not persuasion.

> If the drivers weren't part of that, their quality would lag WAY behind.
> (And to be honest, you do see that in several binary drivers: bad
> quality. Nvidia seems to be improving somewhat there but they didn't
> always have that; others are still struggling).

Hm well, can't speak for others but the *only* problems I have with the
nVidia drivers are when kernel upgrades break them. Last time it was 4k
stacks, this time it's apparently udev.
 
> btw the glibc NSS abi is changing all the time ...  X is about to change
> their driver interfaces because they suffered too much bitrot.

Last time I looked in the X team were bending over backwards to avoid
breaking the binary driver ABI. The huge changes to the rendering model
brought by the COMPOSITE extension may in the end mean breaking backwards
compatibility but there is a huge difference between breaking stuff just
to rename a function and breaking stuff due to a core rendering model
change.

> they do have a choice. they can do things on the other side of the
> designed stable interface. For example, NVidia could do this. 

I'm pretty sure they could not do this without an unacceptable
performance loss. Perhaps an nVidia engineer should be asked, they are
easy to contact.

thanks -mike