Slashdot Mirror
3DLabs Releases Linux Drivers
wilfie writes "3DLabs have released linux drivers for their for Wildcat III and Wildcat 4 Graphics accelerators. Being closed source they'll taint your kernel, but what the heck. Press release with penguin-friendly quotes available too." DataSquid has a note about ATI's Linux support: "While on the job hunt, I came across this posting at ATI seeking a project team lead. Last on the list of key responsibilities is "Act as a leader to improve the overall quality of Linux support at ATI." Good news? Certainly better news than what was suggested before."
The point, of course, is that when you post on the lkml saying "Wah wah my kernel's dead", they can come back and say "Sorry - we can't fix that because you're running code in your kernel which we don't have access to". Or possibly something less polite :-) But anyway, that's why the taint flag exists.
Paranoia isn't an infectious condition, it's a way of life
We have that already. Older ATI chipsets (like Rage 128) have decent support from the DRI project (http://dri.sourceforge.net) and give quite acceptable 2D and hardware-accelerated 3D performance for most uses, at a quite reasonable cost. DRI supports some newer ATI cards as well. However, it can't support chipsets whose makers won't release specs, such as those made by NVidia.
Nonaggression works!
Consumer-class video cards push fillrate, which are ideal for games. They suck for 3D professional work, though.
Workstation-class video cards, which 3DLabs make, push polygons over fillrate. They suck at Quake 3, but are great for 3D professional work.
The older Matrox cards are already very well understood, documented and supported by Matrox. Open Source drivers for loads of systems abound. Same for the older ATI chipsets.
Another point of the taint flag is that you cannot make a binary distribution of a tainted kernel. Under the GPL, if you distribute any modifications of the original kernel, you must include the source, which you can't do if you have closed-source drivers.
I hereby place the above post in the public domain.
In nVidia's case they licensed code from SGI or some company like that. If they GPL'd the code their would be a lawsuit which I don't think they want. Now you could atleast release the spec for the cards , so the community could create its own drivers. That is proabably why many companies don't get involved in Linux for drivers, because the community already made them.
is will their best card with their driver burn
smoking rings around my nvidia card with nvidia's
drivers the way that nvidia cards with nvidia
drivers burn smoking rings around ATI cards with
ati drivers under linux? I tell you this... If
ID does a 180 and doesn't release a Linux version
of Doom3 at the same time, or a reasonably short
time after the windows version, I'm getting
whatever card is fastest under XP and giving up
on Linux gaming altogether. It's really nice
having a quake3 link in my blackbox menu, but
I'm already sick of having to reboot to play
counterstrike, NOLF2, and other games.
For every annoying gentoo user, are three even more annoying anti-gentoo crybabies. Take Yosh from #Gimp for example.
Well, there's the venerable VESA2.0 standard which is implemented in all video cards produced since mid-late '90's, and you can take advantage of this by the vesa XFree86 driver.
The problem is that the VESA standard interface gives you just a framebuffer: there's no 2D or 3D acceleration and that is a real performance hit. (BTW, memory size has nothing to do with speed).
So, what you ask for is already done: you can use any videocard to display a (somewhat slow) X session.
But 2D is a solved problem: if you want that, you don't need an ATI 9700Pro or a GeForceFx, an old Matrox card will do it (the G100, G200 and G400 have very decent drivers, they are fast and the picture quality is superb). People want screaming fast 3D with all the latest features, and while there is a cutthroat race for this among the videocard manufacturers I don't expect really open source drivers.
(Oh, and there's the good old Macrovision stuff for TV outs, which is often cited as a reason why is it, that there are no open source support from the vendors for the TV out)
Real life is overrated.
They fixed their Windows drivers, so let's hope they can do the same for their Linux drivers. ATi's cards are fairly nice; certainly a little better than the dim-looking GeForce FX line, and I say that as someone who is completely neutral when it comes to corporations. (No reason to make generalizations about anything except their products. Unless they're Rambus, in which case they're evil.)
Amusingly enough, the only nVidia card I'd recommend buying right now is their Titanium 4200, the very card that taught them a valuable lesson about market segmentation. In case you missed the whole thing, the Ti4200, Ti4400, and Ti4600 were spaced about $50 apart, except they had a performance difference of maybe 2-4% between each of them. The Ti4400 cannabalized the Ti4600's sales, then the Ti4200 cannabalized the Ti4400's sales. (ATi learned this too with their Radeon 9500 Pro, a fast card selling for less than the Radeon 9700 despite the small speed difference. They were quicker to react than nVidia, though, and stopped the 9500 Pro's production run short. Now they make the Radeon 9600 Pro, which is considerably slower than the old 9500 Pro.)
In the USA, corporations and organizations are treated as singular.
In Britain and many other parts of the world, they are treated
as plural.
For examples of this, try looking at BBC news items.
*sigh* back to work...
In a related note, S3 graphics has released open source drivers for the S3 Savage graphics family. See eg. here.
Well, the same reason as to why we don't have room-temp (or only requiring passive cooling) 500mhz processors for $25, silent single platter 10gb HDs for $25, 256mb 266mhz DDR RAM for $25.
600MHz passive-cooled processor for $15
Silent single-platter 20gb HD for $39 (can't hit the magic $25 price point, but it is 20gb)
256mb 266MHz DDR RAM for $22
It CAN be done for the prices you quote; but your point still is valid. The flashy stuff sells and gets the PR.
Guess you haven't tried RedHat out in quite a while. The telnet server hasn't ever even been installed by default, let alone turned on since the RedHat 6.2 days. Sendmail is installed and turned on by default, but it is only bound to 127.0.0.1, so you can't even connect to it remotely unless you explicitly turn it on.
Hardly. RedHat out of the box in workstation install has no services running by default (except ntp), and the default firewall config only allows in ssh anyway. Even in custom or developer install, only ssh is on.