AMD to open-source ATI drivers?

[Demirug]

It will be interesting to see what the reaction to the open-source Intel drivers will be. If that's a generally positive experience, it will become much easier to see open-sourcing to other graphics companies, although I doubt Intel have anything like the software investment that ATI and nvidia have put in.

From my point of view the Intel driver is worthless as long as you have not the same chip documentation that the guys who have written it seems to have. There are too many “magic code partsâ€￾ that are hard to understand. From the comments I even have the felling the developers sometimes understand it either and have written it only because it was somewhere in the chip documentation. Maybe you can fix some buffer overruns or such errors but if you want to improve the driver you are lost without the right documentation. It’s like trying to learn a language with only a newspaper in this language.

 

[DemoCoder]

Linux/FreeBSD/et al became stable and polished well before the corporate money rolled in. Especially *BSD, because the system is very old. Emacs, for example, had what, 20 years to mature? The money came later. I was using NetBSD on my desktop and for my servers in the early 90s before most monied interests had even heard of the internet or open-source and started jumping on the band-wagon.

I think open-source works best when the core of the app is written by a small closed team, and then it is ported and bug-patched by the community. You just can't get the same results by running an internal issue tracker and have people send in feature requests and bug reports vs having them port, add features, and bug fix directly and then submit patches to you.

 

[Dio]

Linux/FreeBSD/et al became stable and polished well before the corporate money rolled in. Especially *BSD, because the system is very old. Emacs, for example, had what, 20 years to mature? The money came later.

I did say 'Or some other degree of outside funding' - BSD and probably Emacs both had substantial support provided by university staff and students - they weren't primarily developed by code hackers working on their own time, and the FSF funded the development of GNU. In all cases I'd be pretty certain the majority of the code at just about any stage of their development has been paid for by somebody.

Linux is certainly an exception in its very early days. I don't know how far it went before it was picked up by paid staff.