3dfx Glide and DRI Open Sourced 152
jazzman45 writes "3dfx has released glide v3 as open source. There's the link to the driver's page. It has support for XF86 v4 and it's DRI structure! I found a link to someone's screenshots of Q3Arena in Linux. "
Re:How open are these, really? (Score:1)
Dammit, I'm real tired of people saying this about nVidia. Try these [nvidia.com] on for size.
What exactly is the point though? (Score:1)
Perhaps it's for that nice open source feeling
Jeremy
Re:Arrgghhh!!! (Score:1)
Um...that Mesa support for Voodoo2 etc. is just a wrapper around the Glide libraries, which with this new development don't look like they're going to be updated for much longer.
Re:Details? (Score:2)
- |Daryll
Re:Darryls post to 3dfx.glide.linux yesterday (Score:1)
Re:Stupid question? (Score:1)
Everything has changed. The license seems to be an attempt at being compatable with the XFree86 license.
Pan
Woohoo! (Score:2)
Dammit; I keep trying to resist the urge to buy a Voodoo3 instead of waiting for V4, and then something like this has to come along and tempt me again. I suppose I'll be able to hold out till the PPC port is working. Here's hoping that's not till after the V4 is out
Did I make a mistake? (Score:1)
Now, the 3D in windows is wonderful, and the glasses are impressive, but seeing how I'm in linux 99.9% of the time, this doesn't matter a great deal. It seems that performance on a TNT2 in linux is still substandard (I actually haven't gotten it to work at all yet, but that's my own fault), so, with this receint development, would it be in my best intrest to return it and get a Voodoo 3? Also, on a side note, is there a 'TNT under linux' resource that's any better than Nvidia's site?
Re:Arrgghhh!!! (Score:1)
Craptacular (Score:1)
Re:Stupid question? (Score:1)
Here's what happened. 3dfx offered a prerelease of XFree86 4.0 with DRI support for Mesa using glide3x. All of this is new. XFree86 4.0 and DRI allow better support for Mesa rendering. Glide3x is a newer version of the Glide library. Previously only Glide2x was availiable for Linux. Glide3x is a bit different from Glide2x and they are not binarily compatible (like libc5 and glibc2.0 and glibc2.1 are not compatbile). They also aren't source compatbile but a porting job can usually be done pretty quickly.
Now for the open source. Within the same prerelease 3dfx also released the complete Glide3x source code for Voodoo3. This can be built for DRI or as a standalone API. What happened after this source release? Not much yet, its only been availiable for a few days and it took some time for people to notice it.
Joseph Kain
3DFX Cards (Score:1)
I'm sorry if I seem ungrateful and all that, but the tying of the Linux 3D market to 3Dfx cards (as it effectively is at the moment) is a bad thing, we're suffering from relatively low quality graphics compared to the 'doze world...
However, if this means that it's now going to be easier to write glide wrappers for other cards which do support some kind of extra acceleration, then this makes me happy...
Any news on the state of play with Linux and the GeForce yet?
3dfx: yaaaaaaaaaaaay (Score:1)
If you think you know what the hell is really going on you're probably full of shit.
Re:Yet another source license (Score:2)
is restricted in certain countries either by patents or by copyrighted
interfaces, the original copyright holder who places the Program under
this License may add an explicit geographical distribution limitation
excluding those countries, so that distribution is permitted only in
or among countries not thus excluded. In such case, this License
incorporates the limitation as if written in the body of this License.
Funny this looks remarkably like Section 8 of the GPL [fsf.org]
This is not to say that there aren't patent problems. It would be better if they explicitly licenced those particular patents to open source software or at least derivatives of this software. That should make any derivatives redistributable. If they did not do this and decided to sue later for patent infringement, I should think that one would have a case that they acted in bad faith, but IANAL.
If they do not own the patents then all bets are off.
--
Re:Does this mean we can have Glide with a TNT car (Score:3)
It's really just a D3D wrapper, but it may be what forced 3dfx's hand.
---
I guess this is a Good Thing, but.. (Score:1)
-Warren
rendering clusters? (Score:2)
As far as I understand, all the drivers developped nowadays are for interactive 3D or real time 3D. Sometimes in scientific imaging, all you are really interested in is non-interactive rendering, like the stuff you would get from pov, renderman or radiance...
In this respect, it'd be much more interesting to use the 3D accelerator for just console-based rendering.
I know there was a plug for running mesa in parallel that would also support voodoo 1 at some point...
Is there any work done for accelerating this kind of "console" rendering or is all the work these days done on xfree drivers?
---
Woohoo! (Score:1)
Oh yeah, 1st post.
Deosyne
Re:bout the screenshots (Score:1)
--
Re:Has everyone forgot what a tarball is? (Score:1)
Wohhhh!!! Imagine playing Quake on one of these!!! (Score:1)
Oh... wait a sec.. yeah, I guess that isn't actually that off topic for this story... Sorry All!!
Oh well... First Post, then? :-)
--Donate food by clicking: www.thehungersite.com [thehungersite.com]
Mitigated good news (Score:1)
This does not quite convince me that my next 3D graphics board should be a Voodoo-whatever, though. Does anybody know if TNT drivers have been open-sourced yet?
Re:will AGP VooDoo3 work? (Score:1)
Way cool but... (Score:3)
I'd like to see the source for the complete Glide library so that my "old" Voodoo2 can also use Glide3X. But, 3Dfx probably isn't going to release this, and they are equally less likely to update Glide for Linux (at least, they haven't done so in ages - Daryll, any takes on this one?)
Oh well, I planned to go out and buy another 3D card for Q3Arena anyway...
--Fritti
Re:Arrgghhh!!! (Score:1)
I agree - you are a fool.
>Waah waah waahh... Look, 3dfx is still on probation in my book too, but god are you ever an ingrate. Has the Linux Way become to bitch at companies who don't do ALL your work for you?
What exactly is your point? Glide is proprietary, and 3Dfx have been very vigorous in defending it (recall the Glide wrappers for TNT cards), preventing anybody from finding a way around the restrictions on it (like lack of cross-platform support, for a start). I paid exactly the same amount of cash for my Voodoo2 card as any Windows weenie, and yet the company that made that card still insists on treating me like a second-class citizen. Does that make me an ingrate? Or did you just feel like spouting bullshit?
Why don't you take some time out to grow up.
Re:Mitigated good news (Score:1)
now that 3dfx has finally opened up their 3d hardware, ATI is probably feeling pretty lonely.
Now we wait anxiously for XFree86 4.0 and for pure 3d goodness.
Lately... (Score:1)
Re:Way cool but... (Score:1)
man, that sucks
It's good to see all the competition in 3d. Hopefully, market pressure and competition will help get us open source drivers.
bout the screenshots (Score:1)
Re:Sucide (Score:1)
Deosyne
I wonder if.. (Score:2)
Re:Mitigated good news (Score:1)
Re:Thank you 3DFX (Score:1)
Re:Does this mean we can have Glide with a TNT car (Score:1)
Can this be used by non-Creative labs TNT cards, such as the diamond ones?
Anyone any idea? I looked for a proper GLiDE- wrapper for ages, but never could find anything beyond Ultra-HLE support.
http://www.precisioninsight.com/ (Score:1)
They have some deal with ATI for open source drivers on some of their cards.
Snowball fight, anyone? (Score:1)
Re:Performance. (Score:1)
Stupid question? (Score:3)
----------------
"Great spirits have always encountered violent opposition from mediocre minds." - Albert Einstein
Re:rendering clusters? (Score:1)
At root 3d cards liberally approximate lightining and reflection and other things to allow the operations to be performed extremly fast on specialized hardware.
What you are talking about is accelerating raytracing/radiance type imaging which does not make such simple approximations and hence at root just requires a bunch of computational power (hence why titanic was rendered on hundreds of linux boxes and not "rendering cards").
PROBLEMS (Score:1)
Re:rendering clusters? (Score:1)
One word: pbuffer [sgi.com]. It's an SGI GLX extension, but it would be really cool to have...
Long live OpenGL!
Re:PROBLEMS (Score:1)
"The importance of using technology in the right way has never been more clear." [microsoft.com]
Re:rendering clusters? (Score:1)
As always on the internet, this has to be taken with a bit of caution, I don't even know if this is for real or not... but you get the idea.
Isn't there anything on a 3D accelerator that could be used to accelerate pov and the like?
---
Re:Performance. (Score:1)
Re:rendering clusters? (Score:1)
Yes, but this could be used for generating so called low quality animations really fast (keyworkd here: non-interactive)... the new G400 MAXX has some very interesting bitmap bumping capabilities that could be implemented as OpenGL extensions (now it's just available in DirectSomething mode). Or you could use that for medical imaging. Network banwidth would be a problem, but I think it's worth considering...
Re:PROBLEMS (Score:1)
Last I checked they had voodoo drivers (I think for voodoo3, not sure about voodoo2)
They have X drivers as well as FullScreen 3D stuff.
a quote comes to mind (Score:1)
-- i just made this up
"The importance of using technology in the right way has never been more clear." [microsoft.com]
Re:Woohoo! (Score:1)
instead of
-Ben.
Re:Has everyone forgot what a tarball is? (Score:1)
Just use rpm2targz.....
Re:Details? (Score:1)
Re:Way cool but... (Score:1)
--
Re:rendering clusters? (Score:2)
BUT! It's okay when nothing complicated is involved.. as in just the data m'am. In my case that would be the bladder/prostate/rectum compound, no texture, 1 or 2 lights, no shadow... :)
If I were to add the patient skin with realistic semi-transparence and wanted to do functional animation as well, things bet much messier, and I (would) do that in pov... The problem then becomes, okay, everybody does it with clusters, but why? is it because 3D accelerated hardware is looked at the wrong way, or is it because this is not really the main focus (a.k.a games!
A card designed for med imaging would need to do clever things as marching cubes maybe, and also raytracing from voxels (any other suggestions? :)
A side note: for hardware accelerated opengl and computation, I use IDL [rsinc.com] on NT but am waiting for a comparable solution in Linux (yep, sure do!)
---
Is this true??? (Score:1)
However, maybe it's time for the XF86 developers to open up more. I know that the 3.9.xx betas were supposed to start that process up.
Meta Post of Answers (Score:5)
This is a prerelease of the DRI software for 3dfx. It includes FULL source. Because it is a prerelease we know there are bugs and issues. We're collecting feedback on the newsgroup news://news.3dfx.com/3dfx.linux.glide.
It works with the Voodoo Banshee and Voodoo3. Early cards will never support it, because they are 3D only cards. It doesn't make sense to do X on them. (The Rush is an exception, but it is the so degenerate that it isn't worth the trouble.)
The old cards will still work just fine with the old Mesa and Glide. Applications will be linked against Mesa, and if you have an old card it will use Glide2x and be fullscreen. If you have a new card, it will use Mesa+DRI and glide3x and possibly run in a window.
This is a phased rollout. Right now we're in the "stealth" phase. :-) There will be more materials coming out over time and some more press releases to talk about it. So when you see the press releases you'll know about everything coming out.
Why do you care about this? This lets you run multiple apps at once in a window. Performance is just about the same as fullscreen was. So, if all you cared about was full screen quake, this doesn't make any difference. If you wanted to run other apps, this is a big win. It is also the first full DRI solution, which should be helpful to other projects.
This work emphasizes OpenGL. That's why the glide3x defaults to DRI only use. The code to make Glide3x work fullscreen non-DRI is included in the source tree. We want companies to use OpenGL. We realize there's a problem with Glide2x only applications. They won't work in this prerelease. Fixing that correctly means making Glide2x a DRI client. We're working on a solution to do that.
Yes, it is yet another license. (I had nothing to do with that) If you have specific problems with the license bring them up on the newsgroup. They may get changed.
What about other distros/OSes? You've got the source, go for it. We're still doing a prerelease. We've got other problems to worry about first.
All this work was done by Precision Insight and 3dfx.
I think I've got all the relevant questions. Hopefully this will get moderated up. I'll look for other questions later.
- |Daryll
Re:Details? (Score:2)
- |Daryll
X not designed for games (Score:2)
Realism can also come from using advanced features on the card which let gl programs do more work for little cost in frame rate. Open sourcing the drivers also makes this possible, but getting the best performance requires programmers to know which trade offs to make on which card. Although this is possible don't expect it too soon.
This is news because a lot of people have these cards and it pressures other 3d card makers to follow.
Re:Arrgghhh!!! (Score:2)
Voodoo Rush is dead. It was still-born. Take it from someone who was foolish enough to actually buy one at one time.
Waah waah waahh... Look, 3dfx is still on probation in my book too, but god are you ever an ingrate. Has the Linux Way become to bitch at companies who don't do ALL your work for you?
screenshot mirror soon... (Score:1)
F'in cool, hehe!!
MULLY
Could someone enlighten us? (Score:1)
Re:Thank you 3DFX (Score:1)
Re:Way cool but... (Score:1)
Hell, man, you can play Quake 3 on a Voodoo 1 (ONE!) in Linux. I know -- I tried it. Lag was awful, but it was worth it to see the look on my friend's face when he realized I was using the same Voodoo1 that he couldn't get to work with Q3 under Windows...
Like Open Hardware? (Score:1)
Ah, You might mean: Are those Open Hardware? [openhardware.org]
I guess not, at least there is not any common video card in Open Hardware Catalog. [openhardware.org]
Do manufacturers even know that Certification exists, do they even care? ... like GPL is now?
- If not, it's because nobody asks that certification.
Would you like to use it as definition of openess
- Then we must start using it, then others start asking about it ("Huh, what's that?"), and perhaps it can be become well known denifition.
It's up to us
Re:will AGP VooDoo3 work? (Score:1)
Re:PROBLEMS (Score:1)
that's the Control Key, The Alt Key and then the Plus key
Look first, then complain. (Score:2)
DRI.spec
DRI.tgz
glu.tgz
glut.tgz
misc.tgz
unique.patch
[sandeen@Lager sandeen]$ rpm -qpl Glide_V3-DRI-3.10-2.src.rpm
3dfx.gif
Glide3.10.tar.gz
Glide_V3-DRI-3.10.spec
Ooh! Look! Tarballs inside the RPM!
----
Re:I guess this is a Good Thing, but.. (Score:1)
Details? (Score:1)
----
Re:Look first, then complain. (Score:1)
I was pointing out that there *were* source tarballs, you just had to extract them from the RPM.
----
Re:3.9.16 a mistake! -- Is XF86 4.0 dead? No. (Score:2)
Q: I hear a rumour that XiG has bought away several of the XF86 developers.
A: A totally erroneous rumor.
Q: Why no new snapshots?
A: Because the developer in charge of snapshots hasn't uploaded any, probably due to having a life outside of XF86 development.
Q: Has the CVS been shut down?
A: No, in fact an update was made last week.
Q: When will XF86 4.0 be released?
A: Unknown, but it has not been terminated. There are still issues to be resolved and, as with all development, they want to avoid releasing a bad version.
Re:3.9.16 a mistake! -- Is XF86 4.0 dead? No. (Score:1)
*sigh*
matt
Re: (Score:1)
Re:PROBLEMS (Score:1)
What you want to do is to check out Darryl Strauss's excellent work with the with the VooDoo 3 (& Banshee) XServer. Until you install this you will not see any better than 320x200 on your card. Darryl's site is at www.linux3d.org [linux3d.org], but the status of the XServer development is at www.linux3d.org/status.html [linux3d.org]. Be sure to read whatever you find there, there's lots of good information, and you'll need it all. After you get X working correctly, for instance, you'll want to install the Glide RPM's to get your 3d working...
If you're as new to Linux as you're saying, you may want to consider waiting a while before playing with the brand new versions of the 3DFX drivers that are being announced in this /. article, at least until they stabilize a bit. Finally, after you've R all TFM you can find, if you're still having problems and need assistance getting things working, you can point your favorite news reading client at news.3dfx.com and look for the various Voodoo Linux newsgroups. (I don't have the exact names with me at work, sorry.) The folks there are typically very responsive and helpful, but, please, read the last few weeks worth of posts to see if your question has already been answered before posting yourself.
(Hmmmm.... just noticed that www.linux3d.org doesn't seem to be responding right now. A little busy with the big news, perhaps?)
Good Luck!
Re:Does this mean we can have Glide with a TNT car (Score:1)
Thank you 3DFX (Score:1)
Re:Way cool but... (Score:2)
ramblerambleramblewhyamihearpostingstuffon/.
3dfx goes open (Score:2)
Format Fanatic (Score:1)
I haven't noticed a quality difference in PNG, but then again, I've neever been an imaging purist.
Re:rendering clusters? (Score:1)
Well, there is people working on taking advantage of 3d hardware acceleration for computation, but for know, it's pretty much SGI related, because many things that could be called "high" end on the usual Intel world, is standard on SGIs, so, from the SGI user point of view, some of these things can be called "comodity" hardware (really fast buses, texture memory, and the extensions to support and take advantage of those)
IDL is available for Linux. Real support for 3d hardware acceleration is on its way... /me is looking forward for the GLX for G400 support [openprojects.net], Matrox provides real specs so real drivers can be developed.
Re:What exactly is the point though? (Score:1)
--
Re:Does this mean we can have Glide with a TNT car (Score:1)
Re:Look first, then complain. (Score:1)
Re:Does this mean we can have Glide with a TNT car (Score:1)
Re:Way cool but... (Score:1)
OK, I no longer work there, but there is a released glide3x for v2 right? Why can't you use this?
Re:Woohoo! (Score:1)
And where exactly are the hard lines in a q3 screenshot?
Re:Has everyone forgot what a tarball is? (Score:2)
Despite all that I know several people that have this running on Debian, using alien to convert the packages to debs. I've converted the packages to tgz files using alien as well. I've also built the entire thing from source on Slackware 7.0. Given all that I think this prerelease is in very good shape. The only major problem I've seen so far is that it won't run on Suse (The provided binaries or a build on Suse) I'll be looking into that today.
Joseph Kain
Re:What exactly is the point though? (Score:1)
Joseph Kain
Re:Performance. (Score:1)
Re:Woohoo! (Score:1)
Slashdot should get it together and not post headlines confusing part of glide with the whole library. This does not mean you can take full advantage of your Voodoo card with free software - not even close. In order for that to a happen, all the device drivers and the complete glide library would need to be freed.
One step in the right direction, but 3dfx still has a long way to go before they respect the rights of users in having source code for all of it.
Until 3fx gets a clue, it's still no good guys...
Re:Has everyone forgot what a tarball is? (Score:1)
Performance. (Score:1)
Can any non-3dfx card do 3d in a window with linux? Faster?
Does this mean we can have Glide with a TNT card? (Score:2)
I'm talking more about Windows than Linux here. Normally I hate Windows just as much as (if not more than) your average Linux user does, but gaming is one area where Linux just can't compete with Windows. This is not flamebait, its a legitamite question, even in a crowd of Linux users. Remember, open source can apply to Windows stuff, even if it usually doesn't.
Re:3dfx goes open (Score:1)
(This is only a thought, and it may or may not reflect the truth. I do not claim that it does, it is just a thought.)
--
Re:Way cool but... (Score:2)
Folks, this is the first prerelease. Hang in there. There's more to come.
- |Daryll
Re:Woohoo! (Score:2)
The FULL code to glide3x for Voodoo Banshee/Vodoo3 is included. You CAN take advantage of your Voodoo card with free software!
The release is focused on the DRI, so if you do a make at the top of the tree, you get a DRI compatible version. If you go into the sources and change a few symbolic links (foo.c.dri to foo.c.orig) then you get a fully functional full screen version of Glide.
Clue check, you failed...
- |Daryll
Re:Details? (Score:2)
- |Daryll
Re:Woohoo! (Score:2)
Re:Details? (Score:2)
OOPS!
- |Daryll
How open are these, really? (Score:3)
As I still only have a riva128, matrox's (imho) better open-ness will lead me to get a g200/400 when I do upgrade...
David
Better than it sounds (Score:2)
But from the fact that opengl can be
implemented as a thin(ish) layer over glide
which makes it wa-hey cool for voodoo3
owners.
As one previous poster mentioned, it would
be great if 3dfx would release stuff which would
allow the older voodoo cards to work too.
Yet another source license (Score:2)
Here's the preamble:
program interface (API).The license is intended to offer terms similar
to some standard General Public Licenses designed to foster open
standards and unrestricted accessibility to source code. Some of these
licenses require that, as a condition of the license of the software,
any derivative works (that is, new software which is a work containing
the original program or a portion of it) must be available for general
use, without restriction other than for a minor transfer fee, and that
the source code for such derivative works must likewise be made
available. The only restriction is that such derivative works must be
subject to the same General Public License terms as the original work.
This 3dfx GLIDE Source Code General Public License differs from the
standard licenses of this type in that it does not require the entire
derivative work to be made available under the terms of this license
nor is the recipient required to make available the source code for
the entire derivative work. Rather, the license is limited to only the
identifiable portion of the derivative work that is derived from the
licensed software. The precise terms and conditions for copying,
distribution and modification follow.
I noticed two obvious hitches:
section 3.2(f) says:
You do not make any use of the GLIDE trademark without the prior written permission of 3dfx.
And yet, a la GPL, you must insure that all recipients receive a copy or be referred to "this License" which is defined as the "3dfx GLIDE Source Code General Public License". Hmmm.
I'd guess this is an oversight, since they seem to be attempting to hold to the spirit of the GPL.
Also, section 4.2 says:
If the distribution and/or use of the Program or Derivative Works
is restricted in certain countries either by patents or by copyrighted
interfaces, the original copyright holder who places the Program under
this License may add an explicit geographical distribution limitation
excluding those countries, so that distribution is permitted only in
or among countries not thus excluded. In such case, this License
incorporates the limitation as if written in the body of this License.
This seems to say they can take away one's right to redistribute at any time, especially since the readme lists a number of US patent numbers. This seems like a real show-stopper.
Comments?