[conspire] Linux compatibility ??

[Daniel Gimpelevich]

On Sun, 24 Aug 2008 20:36:37 -0700, Rick Moen wrote:

> A commenter on a Web forum says:
> 
>   Sorry, no hardware acceleration for the VIA Chrome 9. I'm in contact
>   with one of the key developers for X drivers for VIA chipsets (VIA's
>   most infamous detractor - libv) and aware of the horrors he has to go
>   through in order to get anything functional (VIA's 'support' in this
>   area mainly consisting of random and sporadic drops of binary, not
>   source code, which he then has to painstakingly reverse engineer in
>   order to get something functional).

That's only part of the story. Luc Verhaegen (libv/dwdeath), the 
maintainer of the http://unichrome.sf.net project, is one of the people 
Novell is paying to develop the AMD/ATI drivers for current hardware in 
SuSE. As such, the extensive development that proper Chrome support 
really needs is largely in the "someday" pile. VIA's drivers consist of 
acceptably licensed kernel-space code and binary-only user-space code. 
The kernel-space code consists of hard-to-read patches to the direct 
rendering code maintained by Thomas Hellström, and the user-space code 
consists of insanely buggy and unstable XOrg modules. Nevertheless, the 
needed reverse-engineering is not what needs the most work; actual 
implementation is, and adding more cooks to the broth isn't very 
feasible, due in part to VIA's idea of an NDA. Nevertheless, the video 
driver situation with Chrome9 is somewhat more functional than the HD3200 
(in the other laptop mentioned) in any currently-available distribution 
release.