[conspire] video drivers (was Re: "madwifi" is proprietary sludge (was: driver))

[Rick Moen]

Quoting Daniel Gimpelevich (daniel at gimpelevich.san-francisco.ca.us):

> The framerate freaks and gamers unanimously eschew ATI in favor of nVidia,
> and typically give three major reasons: 1) ATI proprietary drivers are
> unusable. 2) ATI open-source drivers require more modest hardware. 3)
> Everything ATI makes constitutes more modest hardware. I say that ATI
> hardware that their proprietary drivers no longer support is a "best of
> all worlds" compromise EXCEPT the framerate freak/gamer world.

The framerate freaks (who _are_ almost entirely gamer kiddies)
unanimously frame the discussion as "NVidia versus ATI".  As apparently
do you.  I'm really sick of such assertions and perceptions clogging up
discussion of Linux video, and a year or two ago decided I'd simply not
put up with it, any more.

It's obvious where that perceptual distortion originated:  It was the
dot.com crash, whereupon just about everybody in computing _but_ gamer
kiddies stopped purchasing hardware, hunkering down and either trying
not to get laid off or skirting the ragged edge of starvation.  

The result?  Both purchasing and discussion of Linux hardware was suddenly
dominated by Win32-oriented gamer punks, contemptuous of open source,
using framerate as their sole criterion, and generally managing to shout
down all other considerations.  For several years, they were taken
seriously because, to a first approximation, they were the only group
spending much money on PCs.  So, we got neon-lit, garish everything,
portable mini-server boxes suitable for transport gaming "parties", and 
a plague of proprietary drivers urged on us by companies like Atheros
who, more often than not, simplied lied and called them "open source".

Anyhow, well, screw that.  No more of that chozzeroi.  Kiddie time is
over.

> >> The Intel [945G) driver isn't quite open-source,
> > 
> > My information suggests you are in error.  Do you have evidence to
> > support your claim?
> 
> How is this license in any way open-source?
> http://downloadmirror.intel.com/df-support/9722/ENG/readme.txt

_That_ isn't, but the X.org intel-agp/i915 driver combination is, and
provides respectable 3D framerates.

One lesson to glean from this experience:  Go to manufacturers for Linux
drivers only as am absolute _last_ resort.  Some of us learned that over
a decade ago.


[Matrox:]

> Then I'm sure you can report on their status regarding good 2D support,
> video capture, multi-head, S-Video and composite out, etc.

2D support is honestly all I give a damn about, at the present time.  

Multi-head:  As you are well aware, I have deliberately gotten rid of
all but one monitor in my household (not counting Cheryl's, which is 
hers and not available to others), the sole remaining (LCD) screen being
almost entirely present merely as an emergency console mechanism for the
house's servers.  By preference, we use laptops as our consoles.  Ergo,
the opportunity for multi-head doesn't arise.

However, Multi-head is reliable said to work fine on X.org mga driver
without Matrox-issued proprietary crapware on G450 and G550.

As you're probably aware, S-Video and composite-out ports are the target
of ongoing Hollywood copyright-baron thuggery, with the video card
manufacturers trying to appease them by making the ports as difficult to
reverse-engineer support for as possible.  Accordingly, suport for those
ports from GATOS-project open source code is reported to be hit-or-miss,
depending on model.  If I personally cared, I'd probably have specifics.

If your point is that open source has enemies in Hollywood and
elsewhere, and that manufacturers have a demonstrated lack of backbone
in that area, those of us who've swamped courtrooms in favour of DeCSS
and freaked out Adobe management by picketing their headquarters
building already knew that, thanks.  But we're in this for the long
term.