Re: Re: whose savage driver is it anyway ?

["Mike A. Harris" <mharris@www.linux.org.uk>]

On Thu, 17 Apr 2003, Nilmoni Deb wrote:

>> We should really allow Tim to maintain his own driver.
>> I bet he would like to do so.
>> Everything else is a mess.
>
>I hope thats what happens eventually, since he's always one step ahead of
>the official version.
>
>Also, I can confirm that the 1.1.27t driver fixes an xvideo bug in the
>previous drivers which made video players display an all-green window at
>color depth 24 on some savage chipsets (for example, savage IX-MV) when
>using the xvideo extension.

1.1.27t fixes many problems users have reported in Red Hat
bugzilla over the 4.2.99.x development period.  While the binary
driver of 1.1.27t was available for a while, the accompanying
source code of 1.1.27t was not available until a while after the 
binary was posted to Tim's site due to a mishap.  I updated the 
savage driver in 4.3.0 to Tim's driver in Red Hat Linux 9, but I 
didn't just replace the 4.3.0 driver outright, I took Tim's 
driver and applied XFree86 CVS /fixes/ on top of that, then 
removed the /fixes/ that weren't really fixes.  ;o)  This driver 
will show up in logs as 1.1.27mh as I wanted to distinguish it 
from Tim's stock driver, and from XFree86.org's stock driver.

There is one known bug in 1.1.27mh with 24bit depth.  I believe 
this bug is also in Tim's 1.1.27t, however not having the 
hardware I can't confirm this yet.  There is a bugzilla entry on 
this already and I'm tracking it down.  It's in virtual NEEDINFO 
right now.

With binary driver snaps being made often now and in the future, 
I believe it would be an asset to have Tim maintain his driver 
in-tree and be able to modify it right in CVS.  It would keep the 
driver more up to date, and also allow a much wider amount of 
community testing than the driver gets currently.

I don't know if Tim is interested in CVS access to the savage 
driver or not, but is this something that could be considered?

Also, Thomas Winishhofer (sp?) maintains the sis driver
out-of-tree.  Similar suggestion of CVS access to the sis driver
for Thomas.

Just to be clear so nobody reads me wrong, these are just some
open suggestions which I think might benefit the future of the
savage and sis drivers, and thus also the the users of the
respective hardware, as well as the XFree86 project.

No expectations, just making a suggestion for consideration.

Thanks,
TTYL

-- 
Mike A. Harris


_______________________________________________
Forum mailing list
Forum@XFree86.Org
http://XFree86.Org/mailman/listinfo/forum