Open nVidia Linux Driver Pledge Nearly Complete

Ciarán Mooney writes to let us know that the Pledgebank drive to raise $10,000 for Project Nouvaeu is almost complete — at this moment it needs only 196 more people to sign up. Project Nouveau aims to provide open source 3D acceleration for nVidia cards. The drive was started by David Nielsen, whose blog explains what he hopes will happen.

Comments

Huh?

[Otter]

People can spend their money as they see fit, but giving $10,000, no strings attached, to a project whose only accomplishment (unless I'm missing something) is "Currently, nothing works" seems like an odd prioritization.

For that matter, why bother with a "pledge drive"? If you think they need $10, why not just send them $10?

Re:Huh?

[eln]

His blog entry says it's basically a $10,000 "thank you" for taking on the project. Seriously, what? I'm giving someone a big pile of cash to thank them for taking on a project, even when they haven't made any meaningful progress toward completion of that project?

I hereby announce I will take on the project of solving world hunger. Please give me a giant no-strings-attached donation as a "thank you" for my initiative. I will then make very little progress toward my goal before finally abandoning it as too difficult.

Re:Huh?

[Anonymous Coward]

a project whose only accomplishment (unless I'm missing something) is "Currently, nothing works"

Yes, you are definitely missing something.

Re:Huh?

[Kalriath]

The project's official website is even more interesting. It explicitly says they have no affiliation with the pledge drive, and don't need money.

Re:Huh?

[Otter]

In fairness, the Nouveau guys (Heh, I hadn't grasped the name before but that's fairly clever...) aren't asking for money, say they don't need the $10K and make it clear that they're not expecting to have a reliable driver for the Fedora 7 release. The hype isn't their fault and I look forward to seeing what they come up with.

Re:Huh?

[runderwo]

Yes, you are missing something, since you are obviously not following the developer blog: 8 8

Re:Huh?

[Citizen of Earth]

I hereby announce I will take on the project of solving world hunger. Please give me a giant no-strings-attached donation as a "thank you" for my initiative. I will then make very little progress toward my goal before finally abandoning it as too difficult.

The U.N. could take a lesson from your honesty.

Re:Huh?

[Anonymous Coward]

Well, I for one would like to encourage such projects and I'd be happy to donate some money. This is not about getting value for your money, it is about showing someone else that you appreciate their effort.

I'd rather give someone money in appreciation of what they have done than to a manufacturer that writes a bad driver and wants to make money on it.

Re:Huh?

[newt0311]

The U.N. could Bzzt... wrong. To my knowledge, the UN is not capable of doing anything. your use of the word *could* implies that you have not yet grasped this vital fact to kep in mind when reffering to the UN.

Re:Huh?

[Compaq_Hater]

True, but why not send that same money to develop a card that is open source and can do the same thing as the Nvidia or Ati cards can 3D wise but useing open spec'ed graphics chips that are designed for Open Gl from the git go.

I know i read some where back in 2004 I think about a company that was attempting to do just that, maybe if they are still doing it talking to them about supporting PCI-E would be possible ?.

CH

Re:Huh?

[Anonymous Coward]

"...they haven't made any meaningful progress..."

Have you ever heard about how the "nv" driver is written in HEX rather then C? One of the first tasks in the nouveau project was to clean up the "nv" code to be more readable.... I wouldn't claim it's finished but that part has atleast made alot of progress already. (Even if the nouveau isn't anywhere close to a drop in replacement for the binary driver, and won't be for many years..)

People have to realise that this is a gigantic project, and it will probably never be "finished" (there will always be more to do). But any progress is better then the current state of the "nv" driver. Please RTFS before claiming _NO_ progress has been made. //fatal

Re:Huh?

[Anonymous Coward]

I'm sure someone at the UN or a beauty pageant might be able to help you.

Re:Huh?

[kanani]

so did you donate?

Re:Huh?

[Littleman_TAMU]

hardware != software There are a lot of smart hardware developers out there, but with all the decisions you have to make regarding architecture, I doubt anything would get done unless a 'hardware Linus' stepped up. In addition, even if you could collaboratively design a competitive chip, you'd then have to produce it. This, I think, is the biggest problem. There's no way an open source project could get the kind of manufacturing support needed for a viable project; and if you tried to produce them as needed, it'd be extremely expensive. I like the idea and would love to join a project like that to learn from hardware gurus (I'm new to the field) and contribute to the open source community, but I just don't see it as feasible. As I recall, Nvidia does their new chip development and testing on a bunch of FPGAs. That's a lot of money for an open source project, not to mention the logistical problem of where to store these FPGAs and do the actual testing. Hardware is just a different animal. For software, you just need a cheap PC, a compiler, and a text editor (or dev environment + source control). There might be some OSS versions of board layout software or FPGA development environments, but I don't know of any. I'd be very surprised if anyone's actually started this type of project, but I'd welcome it and try to help.

Re:Huh?

[Anonymous Coward]

Especially since the official NVidia released binary drivers "Just Work".

This is a worthy cause

[MountainMan101]

With new technology like AIGLX, XGL and XEGL emerging, having open source drivers for 3d cards is very important. Along with the recent R300 work for the ATI cards, this will bring much improved graphics to the Linux Desktop regardless of architecture. I only hope that the ATI X200M card gets open source support soon too (obviously not from nouveau).

Also Fedora 7 (dure April) intends to include the nouveau drivers - which is great as out-of-the-box Fedora can't include the binary nVidia driver necessary to have AIGLX working.

And to anyone who thinks this is unnecessary as there is the binary driver - just wait until you card is dropped from the official support and the old driver stops working with some future kernel.

Re:This is a worthy cause

[drinkypoo]

With new technology like AIGLX, XGL and XEGL emerging, having open source drivers for 3d cards is very important.

While I agree with this statement, I think this project is the wrong way to go about it, simply because we do finally have a vendor who has committed to open source driver support: Intel. Now, I will grant you that their cards are slow and crappy but they should be up to the task of accelerating the linux desktop. Also, the current release supports only an integrated video chipset and some older cards... but voting with your dollars is an absolute necessity. For any non-gamer, it should be a sufficiently powerful graphics system, and the G965 Express Chipset supports Core 2 Duo and Pentium D, so you can combine it with very good CPU power. If I were building a system today (aka if I could afford to build a system today) this is the combination I would elect to use.

But most importantly, we need to monetarily support vendors who give us working hardware with working linux drivers, or even vendors who simply give us enough information to write drivers. This is not ATI or nVidia. This apparently is intel. They're also just about the only vendor providing any useful wifi drivers.

If we actually spend money to sponsor driver development this will be a clear message to all graphics card manufacturers that we will put up with their bad behavior.

Re:This is a worthy cause

[fangorious]

just wait until you card is dropped from the official support and the old driver stops working with some future kernel.

Open source drivers drop support for devices too. And unless you're a kernel module developer, you're just as much at the mercy of others as you are with a binary driver from the manufacturer.

Besides, isn't patent licensing part of the reason nVidia and Ati won't release fully OSS drivers? I believe Intel has patents on certain memory bus related technologies which are used by both nVidia and Ati.

Re:This is a worthy cause

[mauriatm]

"Also Fedora 7 (dure April) intends to include the nouveau drivers - which is great as out-of-the-box Fedora can't include the binary nVidia driver necessary to have AIGLX working."

I'm not sure what level of confidence can be placed in a very incomplete driver that (AFAIK, please correct me) still has not had any releases. As if the driver will be mature enough by April for even a notable minority of users who probably will be better off with the binary driver.

"And to anyone who thinks this is unnecessary as there is the binary driver - just wait until you card is dropped from the official support and the old driver stops working with some future kernel."

The nature of this comment implies old and outdated hardware. So I fail to see how this is any different from developers discontinueing support on hardware they no longer use or find too difficult to support. Merely implying that because a driver is fully open source, it will be supported indefinitely is not realistic.

Although to be fair, I personally would love to see every single driver for all hardware 100% open source. Pipe dreams, I know.

Re:This is a worthy cause

[chromatic]

Besides, isn't patent licensing part of the reason nVidia and Ati won't release fully OSS drivers?

I can't see how, unless someone's somehow managed to obtain patents that don't disclose information publicly and, as such, would suffer material harm in disclosing the patented ideas publicly by releasing source code.

In other words, any vendor that tells you that is lying.

Re:This is a worthy cause

[PCM2]

If we actually spend money to sponsor driver development this will be a clear message to all graphics card manufacturers that we will put up with their bad behavior.

Why? How does your spending money to write open source drivers affect nVidia one bit? Why should it even be on their radar? As far as I can see, it doesn't send any "message" at all, except that there is a small but very vocal minority of users that is willing to spend actual money on products that are compatible with Linux -- but I suspect nVidia already knew that.

Re:This is a worthy cause

[drinkypoo]

Why? How does your spending money to write open source drivers affect nVidia one bit? Why should it even be on their radar?

This shouldn't be that hard to figure out - apparently even the moderators got it this time. See, corporations only feel hits to the wallet. Most of their feedback comes from sales figures, and if they get less love than their competitor (or simply less love than they expect) they hurt, they know something is wrong. Unfortunately, they don't necessarily know why.

However, if ATI or nVidia should lose some market share, they will certainly know that it is not because of their lack of linux support, simply because the OSS community is willing to do the work itself. The proof of this principle is that people are willing to spend money to have someone else do their job for them. Simply buying their products is bad enough, but spending MORE money to support them (they benefit from a driver because it can increase sales) is a clear statement that they don't need to develop open source graphics drivers.

If you really think that this is not on their radar, you are incredibly naive. Linux is the fastest-growing segment in computing, Linux is the only operating system gaining market share in the server space, and Linux is probably the only platform gaining any significant ground in education. Linux will only become more important with time, and Windows less. The change shows every sign of being extremely slow, but that doesn't mean that it's not occurring.

Finally, if it were so unimportant as to not even be on their radar, they wouldn't even have developed their own Linux drivers, closed and crappy as they may be. (Well, nVidia's work pretty well... too bad about ATI.)

Re:This is a worthy cause

[Anonymous Coward]

I think it more likely that the concern is not releasing their patented information (which is of course already released) but the invitation to lawsuits by others holding patents.

Re:This is a worthy cause

[zcat_NZ]

"isn't patent licensing part of the reason nVidia and Ati won't release fully OSS drivers?"

One of the possible issues is _lack_ of patent licensing. Nobody really knows what trivial and obvious techniques have been patented by some patent-troll, but as long as the patent troll can't prove nvidia are doing something the troll's patent potentially covers, they have no reason to sue or shake nvidia down for license fees. Open source drivers would feed the trolls.

Re:This is a worthy cause

[ReachingFarr]

There is also the possibility that nVidia has licensed some patents for their hardware/drivers and they don't actually have the right to disclose some of the code. If they got patent X from some company, which they incorporate into their binary driver, they couldn't open source their driver without removing that piece of code first. Now this wouldn't be a big deal if we are talking about a single function or a few lines of code, but if, for some reason, they have a lot of licensed code they would have to remove it all and we would be left with a very incomplete OSS project.

Re:This is a worthy cause

[epine]

Interesting timing. I have a E6600 system on order with a $50 Asus EAX550 video card based on the hoary ATI R300 core so I could run an open source video driver. Plenty fast enough for 2D and potentially some low-end (non-game) 3D. I tried hard to find something newer and faster, but failed. Matrox has a fully open source driver for some of its older cards, expensive, with lamentable performance, and the second head wouldn't drive the required frequency, which completely negates Matrox's long standing reputation for excellent finals. What I used to like about Matrox is you always knew what you were getting, even if it was a little behind the curve. Then the day came when Microsoft update pushed a new Matrox driver that eliminated multidesk support with narry a "this might screw you over" or "really do this?" I was in the middle of a deadline push and lost half a day discovering that Matrox had fed this into the Microsoft update pipeline in full deliberation. It proved faster to buy an ATI product than research alternative multidesk implementations in software. Still, I have a fondness for what Matrox used to stand for back when NVidia was setting benchmark records with finals that rendered fifty tints of pastel grey.

Since I collected these links just two days ago, I might as well include them:

http://www.skynet.ie/~airlied/talks/ols06/ols2006. odp -- DCC 2006, MIME problem, but opens with evince directly
http://www.skynet.ie/~airlied/talks/ddc05/ddc_pres .sxi -- DCC 2005, didn't read this one myself
http://free3d.org/
http://intellinuxgraphics.org/
http://dri.freedesktop.org/wiki/
http://www.phoronix.com/scan.php?page=article&item =576&num=1
http://www.phoronix.com/scan.php?page=article&item =463&num=1

Re:This is a worthy cause

[TheRaven64]

Open source drivers drop support for devices too. And unless you're a kernel module developer, you're just as much at the mercy of others as you are with a binary driver from the manufacturer. Not quite. If having a maintained driver is valuable to you, then you can offer money to someone to maintain an open driver, even if you don't have the skills to do it yourself. There is also the matter of incentives. On open driver developer has an incentive to keep maintaining the driver as long as they find it useful, while the manufacturer has an incentive to 'encourage' you to upgrade as soon as they release new hardware.

Re:This is a worthy cause

[dfn_deux]

Besides, isn't patent licensing part of the reason nVidia and Ati won't release fully OSS drivers? I believe Intel has patents on certain memory bus related technologies which are used by both nVidia and Ati.

AFAICT (IANAL) this is exactly correct. There needs to be some sort of tipping point reached wherein the projected cost of IP based liability issues is outweighed by potential revenue to be gained through sales based on open drivers. I personally am not too opposed to binary globs with an open ABI; unfortunately I, as well as most linux users (and potential ATI/NVidia customers) have very little say as to what gets accepted into the mainstream kernel.

Re:This is a worthy cause

[cortana]

According to NVIDIA, there is no demand at all for free software drivers for NVIDIA hardware.

Re:This is a worthy cause

[kimvette]

But most importantly, we need to monetarily support vendors who give us working hardware with working linux drivers, or even vendors who simply give us enough information to write drivers. This is not ATI or nVidia. This apparently is intel. They're also just about the only vendor providing any useful wifi drivers.

Yeah, as soon as they offer accelerated graphics with dual DVI I will gladly buy from Intel. Until then, NVidia makes a great, if proprietary solution.

Re:This is a worthy cause

[vandan]

Open source drivers drop support for devices too.

That's not entirely accurate, and not a fair comparison. When a corporation drops support of their product in a binary-only driver, that's the end of the story. When an open-source driver 'drops' support of a product, what they're doing is failing to maintain support. Other people are free to pick up the slack. If a device was supported by open-source software at one point, getting that support up-to-date is far easier than, say, starting from scratch.

And unless you're a kernel module developer, you're just as much at the mercy of others as you are with a binary driver from the manufacturer.

Not really. This story demonstrates that if there is sufficient demand for an open-source driver, it's quite easy to raise enough funds to hire a developer to bring an open-source driver up-to-date ... which again, is a different beast entirely from hiring a developer to start from scratch.

And unless you're a kernel module developer, you're just as much at the mercy of others as you are with a binary driver from the manufacturer.

Not at all. There's a big difference between an excuse and the real reasoning here. What nVidia and ATI are basically saying is that they can't give out documentation on programming the hardware ... register details, setting up DMA, basically talking to the card. This is absolute bullshit. They might be right when they say that they can't open-source their software , because, as they say, some of their software ... ie their OpenGL implementation and other tricks, are legally bound. Fair enough. But no-one is asking them to open-source their drivers. We're asking them to release enough information for developers to write their own drivers. There are no legal issues involved here.

The real reason nVidia and ATI won't release any information for developers is that they see the emergence of high quality, high performance, feature rich, open-source drivers as a threat to their profits. If, for example, Intel can leverage the Xorg / DRI environment ( not to mention developers ), then it means their unit costs are considerably lower than if they have to screw around with writing everything themselves, and sign expensive licensing deals to make use of other people's OpenGL implementation. Intel isn't the only competitor here. Any upstart can make some hardware, hand the open-source community the documentation they need, and guess what ... pretty soon someone has written some drivers for them. And they're fast, feature rich, etc, etc. This is what ATI and nVidia don't want. They have spent millions on their own drivers, and they must protect that investment. There are no legal reasons preventing them from sharing developer documentation with open-source developers.

I believe Intel has patents on certain memory bus related technologies which are used by both nVidia and Ati.

Yes, but that patent covers the hardware, not the interface to talk to the hardware.

Re:This is a worthy cause

[chromatic]

Yes, but that patent covers the hardware, not the interface to talk to the hardware.

Further, that patent is already public information. That's what a patent is.

Re:This is a worthy cause

[Anonymous Coward]

You emphasize the word "proof", and then make claims about Linux being the only OS to gain server or educational market space. Given my firm's recent expansion with Solaris I would like to see your figures.

Furthermore, I would like to see your reasoning behind why nVidia's drivers are "crappy" other than the fact the only source you get is the shim. Based on my personal experience with nVidia's closed-source drivers, they are stable (i.e, zero panics) and provide 3D hardware accelleration to three different platforms, all of which I have experience with this driver on.

Re:This is a worthy cause

[EzInKy]

While I agree with this statement, I think this project is the wrong way to go about it, simply because we do finally have a vendor who has committed to open source driver support: Intel.


I would love to support Intel's efforts, if only they marketed standalone cards. Until then I'll stick with Radeons since I can get OSS drivers which support up to the X850 chipset.

Re:This is a worthy cause

[Anonymous Coward]

According to NVIDIA [robots.org.uk], there is no demand at all for free software drivers for NVIDIA hardware.

I'd have to agree with them. The Linux users who buy nVidia hardware are those who have nothing against buggy proprietary software. The rest of us already bought ATI or Intel hardware with working free software drivers.

(Happily playing Second Life with full 3D acceleration on my Radeon card).

Re:This is a worthy cause

[njchick]

First show me a modern Intel video card I can put into an existing machine.

Re:This is a worthy cause

[Rutulian]

If they got patent X from some company, which they incorporate into their binary driver, they couldn't open source their driver without removing that piece of code first.

Why, exactly, would that be? Where in patent law does it say you can't distribute licensed code? You do know there are mp3 patents, and there are a lot of open source libraries that implement the codec, right?

Just so you know, patent law says you must pay a license to use certain "ideas" (ex: mp3, wma, aac). It does not say anything about whether the resulting code needs to be open or closed source.

Re:This is a worthy cause

[Bert64]

Well said...
I just bought a Macbook instead of a Macbook Pro, as it has an intel chipset instead of the ATI chipset present in the pro. I intend to dual boot OSX and Gentoo on it. I run a few fancy OpenGL screensavers on it, and it handles them just fine, i don't do anything else that would tax it at all.

Re:This is a worthy cause

[PCM2]

If you really think that this is not on their radar, you are incredibly naive. Linux is the fastest-growing segment in computing, Linux is the only operating system gaining market share in the server space, and Linux is probably the only platform gaining any significant ground in education. Linux will only become more important with time, and Windows less. The change shows every sign of being extremely slow, but that doesn't mean that it's not occurring.

Last I heard, nVidia provides drivers for Linux. End of story. If you want to spend your own money to write different drivers, that's your business. But you can't argue that companies (read: the people with money) aren't going to use nVidia cards because nVidia cards don't work on Linux. They do.

Excuse me.

[RightSaidFred99]

Seriously - isn't it somewhat silly to undertake a project of this magnitude (and we're talking a _lot_ of magnitude - take for example redoing drivers for new 8800 line) when it could be instantly obsoleted by one phrase from Nvidia: "OK, nevermind, here are the drivers - we changed out mind."

This sounds, for lack of a better phrase, retarded to me.

Re:Excuse me.

[Aadain2001]

But Nvidia could never legally release their drivers as open source. Many people fail to realize that Nvidia has to license many of the technologies in their drivers from other companies and patent holders (just like ATI, Intel, et al must as well). So, even if the CEO of Nvidia wants to release their driver code as open source, they cannot legally do so without the go-ahead from all the license holders, which will never happen.

Re:Excuse me.

[JonJ]

If they can push nvidia to open either their source code or documentation, then that's great! Right now, getting the driver to work in distros where non-free software is shunned is not much of a hassle, but a little one. If Fedora could be able to ship nvidia(And perhaps ATi)-drivers out-of-the-box then it would be a nice selling point for those that want eye-candy. It would also be easier for those that distribute GNU/Linux to maintain the driver over a longer period of time and perhaps also support older chipsets.

Re:Excuse me.

[businessnerd]

Actually, whether Nouveau successfully releases an open nvidia driver, or nvidia decides to open theirs up, the goal of all of this will have been reached. What the community desperately needs is an open driver for nvidia cards. A large project, like what Nouveau is undertaking, may garner the kind of press necessary to make nvidia change their minds. If nvidia wants any kind of control over what passes as an nvidia driver, it is in their best interests to stop Nouveau by beating them to the punch. You may think that Nouvou had just wasted their time, working on a driver that won't be released, but I see it as Nouveau working on get an open driver released, whether it come from their own developers or nvidia's. It's a win win situation

Re:Excuse me.

[Anonymous Coward]

Nvidia has already stated why they can't make their driver open source: their source code is derived from SGI code (and yes, there was a lawsuit). Their settlement effectively bars them from opening their code.

Re:Excuse me.

[FunkyELF]

People were saying the same thing about Xbox Media Center. When it was in the beginning of development the default skin had it looking like a Windows Media Center Edition PC. I think I heard something about Microsoft releasing some kind of Media Center thing for the Xbox, but guess what....the open source xbmc project is still in active development and is (I assume) far better than the Microsoft one. Even if Microsoft released their Media Center thing as open source, xbmc would still be better. It plays stuff I can't play on my windows machine.

Re:Excuse me.

[hasbeard]

Well RightSaidFred99, Microsoft could announce tomorrow that they're going to open source all of their software under the GPL. But, I don't think I would wait on it. Maybe you have seen something I haven't, but NVidia doesn't seem to be making any noises about opening their drivers.

Re:Excuse me.

[MountainMan101]

Actually Fedora Core 6 shipped with working ATI open source drivers with 3D support (for most cards, sadly not mine). And Fedora 7 will include the nouveau drivers too (according to the release aims).

Re:Excuse me.

[FunkyELF]

Not only that, but I heard that two other reasons for not opening up their drivers are
1. ATI could steal from Nvidia if they aren't already
2. Nvidia sells you a crippled $500 video card for $200 and a hacked driver could turn it back into a $500 card

Re:Excuse me.

[dinivin]

And SGI has already stated that this isn't true.

Re:Excuse me.

[lakeland]

I have to disagree with this.

Yes, Nvidia has NDAs which would be violated if they turned around tomorrow and released a GPL driver. However, those NDAs were negotiated by Nvidia and it would be trivial for them to be renegotiated. I very much doubt the people who developed the components care either way - as illustrated by how quickly intel was able to open-source their driver.

I think the "We'd be breaking our supplier agreements" line is nothing more than a red-herring.

Re:Excuse me.

[dami99]

First, I doubt opening up the drivers really leaves *that* much to be "stolen".

Second, what are the hacks you are talking about

The only one I can recall is one to open the disabled pipelines up on 6800s.... and that's hardly a $300 increase in performance. Not only that, but enabling those pipelines often didn't work or caused artifacts.

Re:Excuse me.

[tinkerghost]

They may not be able to release the code to the drivers as they are - they probably do contain patented/licensed trade secret code. However, they certainly can provide basic - non optimized code to allow interfacing with the chipsets. With that as a basis, the OSS community could certainly work out how to optimize the system - alleviating the trade secret issues, though patents might still be a problem. IANAL, but IIRC, the API's can't be patented, just the code behind them. The rational being that the API's just dictate the interface & there is only 1 way to impliment using the interface - as dictated by the API.

Re:Excuse me.

[schwaang]

1. If they wanted to make the effort, they could release the portions that are not covered by third-party licenses. That would probably help the OSS driver writers a lot even if it's not enough to be a whole driver on its own.

2. If they wanted to make the effort, they could probably get or buy agreement from at least some of their licensors. Again, any bits would help.

But aside from the costs of 1 and 2, nVidia may have other reasons to not open their driver. For example, their lawyers might have made them afraid that they could be held liable for software patent claims against their existing code.

From nVidia's point of view, they probably don't envision the costs outweighing the benefits. So why make that effort?

Re:Excuse me.

[runderwo]

And now you get to provide a link to any documentation about this alleged lawsuit, or alternatively you can explain the existence of NVIDIA's prior obfuscated open source GLX driver released circa 1999.

Re:Excuse me.

[Abcd1234]

Link (from either of you)?

Re:Excuse me.

[Aadain2001]

There is also the issue of "trade secrets" present in the drivers. While you wouldn't think there would be much secrets in a piece of software that acts as a go-between for the OS and hardware, there are more and more cases of drivers actually performing significant "magic" to speed up performance. Releasing that information could give competitors valuable information that could hurt Nvidia. All in all, there are many very good reasons for keeping the driver closed and only a few for opening it up.

Re:Excuse me.

[TheRaven64]

And SGI is worth what? Two, maybe three dollars at the moment? Just buy them and nullify the agreement.

Re:Excuse me.

[mungtor]

Intel has nothing to lose by open sourcing their drivers because their cards suck. They have no interesting technology in that area and really can only gain some market share among linux users.

nVidia and ATI drive the entire graphics card market with their competition and neither wants to give away any info by open sourcing a driver to the very small number of people who even care. Even if you had full 3D accelleration on Linux, there are hardly any games to take advantage of it.

And breaking the supplier agreements is a big deal. Most people here get all bent out of shape when somebody accidentally violates (in their opinion) some obscure and poorly worded clause in the GPL. A license is a license. Either respect all of them or none of them, not just the ones that give you the things you want for free.

Re:Excuse me.

[Anonymous Coward]

The Classpath, GCJ, Kaffe and Harmony guys don't seem to be complaining.

Re:Excuse me.

[Anonymous Coward]

Ahh but you see there are professionals buying cards for Linux and not only that, they buy the most expensive cards in the quadro/firegl ranges. The difference between say a regular radeon and a firegl is generally speaking, besides twice the price, the driver. It's a nice little USD billion dollar/year niche and nVidia and AMD are not likely to want to give it away.

Especially since interest for using Linux in the segments of DCC; CAD and other professional 3d visualization is increasing every year. Games on Linux may or may not be lucrative but consumers are not the only game in town.

And it's not about breaking 3d party code supplier agreements, as numerous other posters have mentioned. If it was, they could easily just release more technical details of registers op-codes and other hardware specific data and specifications.

Re:Excuse me.

[fuzzix]

xbmc ... plays stuff I can't play on my windows machine.

If I'm not mistaken xbmc is based on mplayer. The windows builds are decent enough as long as you grab one of the larger codec packs from the site.

VLC is also pretty damn good.

Re:Excuse me.

[JonJ]

Yeah, I had GLX out-of-the-box with my Radeon 9600, but scrolling with firefox was a world of lag. :)

spelling

[s4ck]

can we run it through spellcheck first? it's Project Nouveau. anywya, who carez..

Re:spelling

[drpimp]

Or use Firefox or browser with spell check built in to post??? 2.0 spell check flags the misspelling for me.

What I hope will happen - by D.N.

[adisakp]

Blog Entry: I hope that a bunch of people on slashdot will give me money. The End.

Re:What I hope will happen - by D.N.

[westlake]

Blog Entry: I hope that a bunch of people on slashdot will give me money. The End

Next Entry: God, how I hate Ramen noodles.

Re:What I hope will happen - by D.N.

[ArsonSmith]

Our team of 4 is almost at our goal of $10,000 USD. We only need 196 more $50 pledges.

What is wrong with the proprietary driver?

[mgemmons]

According to the pledgebank website,

[...]leaving the many users of their videocards on popular UN*X systems such as Linux with only the option of using a 2d only driver or using nvidia' notorious proprietary driver.

What is wrong with using nVidia's drivers for nVidia's cards? Is there some issue with the nVidia 3D driver implementation that would encourage an open-source reverse-engineering effort? What does "notorious proprietary" mean? I'm all for open-source, but this just seems to be OSFOSS (open-source for open-sources sake).

Re:What is wrong with the proprietary driver?

[lolocaust]

If we show that we will accept closed drivers/spec on an open system, we've already lost. Especially with the desktop effects becoming more and more important in modern distros. Also, AFAIK, there are no Nvidia drivers for PPC, and then there are people who could learn about the GPU specifics for the sake of it.

Re:What is wrong with the proprietary driver?

[the_humeister]

One of the problems is that the drivers are x86 only (although there are old and outdated Itanium drivers). Another issue is obsolete video cards. nVidia could one day stop supporting the TNT or GeForce. What do we do then? If there are no open source drivers, we're SOL on updates. If there are open source drivers, then we can make continued improvements when needed.

I switched to a FireGL 8700 (R200-based) for this reason (and it was an upgrade from a GeForce FX 5200). With regards to ATI cards, there are usable and stable open source drivers for all R300-based and lower video cards. Additionally, ATI no longer supports R100-based or lower video cards on Linux. Fortunately, the open source drivers are available to pick up the slack.

Re:What is wrong with the proprietary driver?

[gradedcheese]

potential security problems, for one thing. You should not trust 'binary blob' drivers with your kernel-space, right? you can't inspect the source code...

The 'blob' song actually explains it really well, I think:

http://www.openbsd.org/lyrics.html#39

Re:What is wrong with the proprietary driver?

[amigabill]

What is wrong with using nVidia's drivers for nVidia's cards? Is there some issue with the nVidia 3D driver implementation that would encourage an open-source reverse-engineering effort?

For an x86 box I'm quite happy to use the proprietary driver.

But I also have a PowerPC based system. NVidia's proprietary driver doesn't work on that. Same for others who lack the proprietary supported CPU architecture. And I don't use a Mac either, so if some closed-source thing worked OK enough on Macs but some detail made it not work or buggy on my unusual system I'd still be stuck. I didn't think much of this project either until I realized that. I happen to have a Radeon in mine, as it's better supported, but some people would like to use Nvidia without x86, so I pledged to support for them. I'd be willing to put more than $10 toward it as well, but oh well.

Re:What is wrong with the proprietary driver?

[just_another_sean]

Well when I used nVidia's binary driver on Debian Etch I went through two kernel upgrades and each time I rebooted to begin using the new kernel I was greeted by a console prompt instead of an xdm login screen.

Now for me that wasn't much of a problem. I sighed, logged in as root, found the original installer I downloaded from NVidia, ran it, agreed to the license, pressed continue and was greeted with a message about missing kernel headers. Sighed again, downloaded linux-headers-`uname -r`, reran NVidia installer, etc, etc, ad nauseum every time I update the kernel.

As I said, I know why and how I do this but not everyone does and the whole point of bringing true open source 3d graphics to the desktop for Linux users is so they don't have to learn how or why they need to do this.

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

There is already an open source nvidia 2D driver afaik. This project aims to build a 3D driver for modern applications.

The problems with the non-free driver from Nvidia are:

1. There are old unfixed security holes and bugs that cause other problems

2. Nvidia could drop support for a specific card or even an entire OS at any moment for business reasons

3. An open source driver could be ported to more architectures and operating systems than Nvidia can support in its proprietary driver.

Re:What is wrong with the proprietary driver?

[Bent Mind]

What is wrong with using nVidia's drivers for nVidia's cards?

Just a few things off the top of my head...

nVidia has dropped support for cards older than the GForce4. I have a GForce2 with 64MB and TV tuner that would benefit from this driver.

A while back I was running Hardened Gentoo. When I asked the maintainer why the nVidia driver was masked (blocked), he replied:

... One of the very reasons for using hardened is for increased security protections. The way nvidia wrote the drivers is really crappy as does run time execution. That means it behaves exactly as shellcode does, which is the very thing we are trying to prevent in the first place. Now when that glx (libGL.so) gets installed every single package it that links to it then causes a PaX violation.

I suggest you email the nvidia vendor and request that they stop taking shortcuts in the driver code and release something that's
1) PIC proper [no TEXTREL's]
2) stop using JIT.

Several projects have worked to create versions of xorg or window managers that take advantage of 3D hardware. However, xorg relies on nVidia's driver (with nVidia hardware) for 3D. That code can't be modified.

Finally, my understanding is that the nVidia driver only works with x86 hardware. All of my hardware is x86, so I've never verified this.

Re:What is wrong with the proprietary driver?

[chromatic]

See Pragmatic Questions about Binary-Only Drivers.

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

You can apply your argument to absolutely every open source project. Why bother, you ask? So that the user/computer owner has control of the software he/she uses: the ability to inspect it, change it, improve it, and fix it -- not at the whims of its corporate landlord, but whenever desired.

Re:What is wrong with the proprietary driver?

[kosmosik]

> What is wrong with using nVidia's drivers for nVidia's cards?

F.e. they taint the kernel - if things crash (one of my nvidia cards *did* with some Linux kernel version and their binary blobs) you cannot debug and fix it. Hell kernel developers will tell you to go on /dev/tree since they will not waste time on debuging some closed code with their kernel.

Like it or not this is how Linux philosophy and developement looks - we have (and don't want to) no stable kernel ABI and expect everything (at very least the kernel-space stuff) to be open source. If you don't like it go use like FreeBSD or something.

> Is there some issue with the nVidia 3D driver implementation
> that would encourage an open-source reverse-engineering effort?

Yeah, they are closed, they tie you to one architecture (where are nvidia drivers for PPC?), they break with vesafb and so on. They just limit your freedom. You can imagine situation where you have old nvidia card model FOO but it works for you. Now nvidia decides to cease support for this model in their newest drivers. As kernel developement goes on you will find yourself that newest nvidia binary modules do not support your (perfectly working for you) hardware and the old drivers do not work with new Linux kernels - this limits your freedom in some way.

I would be very happy to use open source drivers for nvidia cards. Imagine just installing a Linux distro and graphics just work out-of-the box.

I *do* think nvidia is quite Linux/OpenSource friendly (but actually I run Intel since it works better for me - no hassle - Just Works) but I think at some point they just will have to release specs of hardware or even better open implementation of their drivers - they will be forced to do so by competition.

Re:What is wrong with the proprietary driver?

[TekPolitik]

What is wrong with using nVidia's drivers for nVidia's cards?

Everything. They're poor quality. Prior to 1.0-8774 they were embarrassingly poor and would often crash X. Now they're just unacceptably poor. If you run multiple X servers on one device the driver will often leave the video in an unusable state from which there is no apparent way to recover. If we had source code we could fix these sorts of problems, but we don't so we can't.

Having the system become unusable because of a bug in the nVidia drivers is a regular occurrence for me. I can't remember the last time the system became unusable because of a bug elsewhere in the system.

What does "notorious proprietary" mean?

Notorious for their bugs, proprietary because you can't look at the source code. The latter leads to the former, because if the source were open those of us who have the skills could diagnose and fix the problems in the environments where they occur. As it is, try reporting the problems to nVidia and if you're lucky enough to get a response it's most likely to be along the lines of "works here".

Re:What is wrong with the proprietary driver?

[mgemmons]

I would like to thank everyone for the informative and (surprisingly) flame-free information about why an open source driver would be a "good thing." Much appreciated!

Re:What is wrong with the proprietary driver?

[Jethro]

First of all, I'm currently using the nvidia drivers on an AMD64, not x86.

Second, if nvidia stop supporting the old cards, don't upgrade the driver. Like they're going to put in improvements for TNT cards anyway?

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

If you don't want to be held to the whims of 'corporate landlords', you probably shouldn't throw them your money in the first place. I can understand the desire and the push for open drivers, but it absolutely boggles my mind how people with such hard-core anti-patent, anti-closed-source, anti-corporate mentalities go out and exercise their superior morals by... spending money on patented closed-source corporate hardware. Don't fucking use nVidia cards if you don't like how they're made.

Re:What is wrong with the proprietary driver?

[Abcd1234]

I think you mean FLOSSFFLOSSS - Free/Libre Open Source Software For Free/Libre Open Source Software's Sake.

Re:What is wrong with the proprietary driver?

[Bert64]

So your content to use the old drivers, with the old version of X, on the old kernel... The drivers still need to be updated to support other aspects of the system that they are tied to.
I have an old FireGL card, for which there are only drivers available for xfree86 4.2.0 on x86, rendering the card rather useless.

Re:What is wrong with the proprietary driver?

[the_humeister]

First of all, I'm currently using the nvidia drivers on an AMD64, not x86.

It's still x86, just 64-bits. That's like saying real mode isn't x86. Well, it is. It's just 16-bits instead of 32-bits (which is what is normally thought of as "x86").

Second, if nvidia stop supporting the old cards, don't upgrade the driver. Like they're going to put in improvements for TNT cards anyway?

OK, but then you upgrade the kernel and your video card stops working because of interface changes. Now what? In addition the binary drivers had a security hole for more than year until it was finally patched late 2006. If the drivers were open source, they would have been patched much earlier.

Re:What is wrong with the proprietary driver?

[T-Ranger]

Then run it in VESA mode. That will work with anything. Oh, you don't get super-acceleration? Well, save your pennies for a month and get a card made this century.

Re:What is wrong with the proprietary driver?

[thopkins]

Why don't you just use the debian package instead of Nvidia's install?!?! It's much easier.

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

I agree with you somewhat, but there are limits to what one can do practically. For the average mortal, you can choose between using closed corporate hardware, secretly designed and manufactured, or using no computer at all (besides perhaps something hacked together from basic components and capable of operating as a simple calculator or text terminal).

We simply don't have the means to manufacture our own open hardware that is competitive. I hope this will change, and there are people making progress in this respect.

As for using closed hardware (nVidia video chips) vs. mostly-closed hardware (Intel video chips), one can make that compromise in either direction -- but the difference in openness from using the non-nVidia hardware is minimal, and it is impossible to write 3d graphics programs to run at a reasonably comparable speed. When this performance is important, I tend to favor the development of open drivers for the nVidia hardware; the completion of such drivers would essentially put that choice on the same ground as the Intel option: open drivers, closed hardware. It's still a compromise, but the next step down (to an entirely open hardware graphics card) is so far down in performance as to be unusable at this point.

I might see where your misunderstanding arises. If you believe that the "anti-patent, anti-closed-source, anti-corporate mentalities" are arbitrary, religious drives, then your misunderstanding makes sense. But for someone who actually holds those mentalities, it is easy to see that they can be held from an enlightened sense of self-interest. Software patents, closed source programs, and corporate control of our computers even after they're purchased: these are all things that explicitly harm the individuals at the branches of the tree. Opposition to them, thus, is natural and logical, but it isn't performed for its own sake. It is performed to eliminate the harmful aspects of them. The goal is an optimal reduction in this third-party control of our individual lives. The path to that goal is not always clear-cut.

Re:What is wrong with the proprietary driver?

[Jethro]

Good point. I didn't consider that an old version would't work with new everything else.

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

Then run it in VESA mode. That will work with anything. Oh, you don't get super-acceleration? Well, save your pennies for a month and get a card made this century.

Your response is the classic reason why we need open-source drivers ... so that we don't have to buy the latest cards when the old hardware still works perfectly well.

Re:What is wrong with the proprietary driver?

[qbwiz]

Maybe the fastest way to do what they do uses a JIT.

Re:What is wrong with the proprietary driver?

[Braino420]

If we show that we will accept closed drivers/spec on an open system, we've already lost.

I don't think Nvidia gives a damn about what drivers are used for their cards... I'll let you in on the secret though, they care about money (!). So, if you show Nvidia that you will buy their cards regardless of the state of their drivers, then you've already lost.

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

So, if you show Nvidia that you will buy their cards regardless of the state of their drivers, then you've already lost.

You're losing the wrong war. It's great to go thumping your chest about Linux users and all, but you forget that the war everyone else is fighting is against the windows marketshare, and guess what, the people who bought computers and ran windows on them aren't going to switch to an operating system that can't run on their system out of the box. "Why does my card need some glue thing compiled? I thought it was already put together?"

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

What is wrong with using nVidia's drivers for nVidia's cards?

What is wrong with using the Dell-supplied Windows XP or Apple-supplied Mac OS? Why run Linux at all?

I'm all for open-source, but Linux just seems like OSFOSS (open-source for open-source's sake).

Re:What is wrong with the proprietary driver?

[Schraegstrichpunkt]

1995 called. They want their ignorance of what free/libre and open-source software is about back.

Re:What is wrong with the proprietary driver?

[Schraegstrichpunkt]

You're not supposed to use the NVidia installer. You're supposed to use module-assistant:

aptitude install module-assistant ; m-a update ; m-a a-i nvidia

Re:What is wrong with the proprietary driver?

[Schraegstrichpunkt]

Maybe it is, so when your machine is compromised, your files will be corrupted faster. Yay!

Seriously, not everybody who wants 3D acceleration is willing to make the same performance-security trade-offs as NVidia's salespeople are.

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

I'm sorry but I don't think this is a problem of being a closed source driver. You would have the same exact problem with an open source driver that didn't come with your distro. Someone needs to develop a hook so that external drivers can be rebuilt whenever the kernel is reinstalled.

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

This is NOT the fault of nVidia, it is the fault of kernel developers who refuse have a reasonably stable ABI.

Re:What is wrong with the proprietary driver?

[lolocaust]

Some people had nvidia cards before they tried out linux. Some retailers mention that there's linux support for these cards, not mentioning to newbs that they have to jump thru a few hoops, just like the atheros based wireless cards. My next computer upgrade will not include an Nvidia, but I do hope that this pledge gets their attention.

Re:What is wrong with the proprietary driver?

[Error27]

Yep... I've installed tons of nvidia cards with linux in the last 4 years and the quality varies. I doubt that 8774 was magical. It may have fixed some cards but there will be new cards soon. The only way is to open source it so that folks can fix the crashing bugs.

Is SLI working yet?

And then there is stupid annoying stuff like this:
http://www.nvnews.net/vbulletin/showthread.php?t=7 4287

Re:What is wrong with the proprietary driver?

[Da Web Guru]

nVidia has dropped support for cards older than the GForce4. I have a GForce2 with 64MB and TV tuner that would benefit from this driver.

My workstation at work (running a GeForce2 MX) is using the 9631 drivers. Granted, I am still running 2.6.18.5, and I don't remember if that driver works with the new kernels (haven't tried a 2.6.19.1 kernel since I upgraded the video card from a TNT2). But, I am not in dire need of any new features from 2.6.19.x as of yet.

Re:What is wrong with the proprietary driver?

[mysticmike]

proprietary means don't work properly with linux

my kernel is tainted so apt won't update it.
the laptop won't suspend so i have to boot up everytime i use it.
then there's that annoying logo.

I am an openGL programmer so i live with a barely working system
because i need the graphics

an open source driver would just work!

would anybody here buy a cpu if the instuction set was a trade secret.

Re:What is wrong with the proprietary driver?

[just_another_sean]

The Debian installer from non-free? Never tried it. It may seem stupid when I just grab NVidia's driver anyway but I don't add apt repositories for non free software. I use non free software, NVidia's driver, VMWare Server, Sun's Java and the Java version of Citrix's ICA client. But I use it as little as possible and when needed... If I am going to use non free software I want to make a concious decision to do so. I don't want this type of software slipping in on me because I pointed to a non-free repository.

So I try to keep the non free programs to a minimum and deal with the hassles associated with them. As soon as possible I replace non free with free. As an example I still use Citrix very occassionally but I convinved my boss at work to let me setup an openvpn server and now connect to work from home through it as much as possible. Haven't used citrix more then a handful of times and then it was from computers I hadn't transferred the necessary SSL stuff to to do VPN. The rest of the above will follow soon, including it seems a 3D enabled NVidia driver!

Re:What is wrong with the proprietary driver?

[Ded Bob]

proprietary means don't work properly with linux

It works well for me under FreeBSD and Linux.

my kernel is tainted so apt won't update it.

That sounds like a bug in apt or the way the driver is packaged; I am able to update it on FreeBSD.

the laptop won't suspend so i have to boot up everytime i use it.

That sounds like a bug in the nvidia driver.

then there's that annoying logo.

Try either of these (from the FreeBSD documentation):
Option "NoLogo" "boolean"
Option "LogoPath" "string"

an open source driver would just work!

You must not use very many drivers. I have run into blocking bugs in open and closed source drivers. This does not address which is actually easier for someone to fix. :)

would anybody here buy a cpu if the instuction set was a trade secret.

Many here have bought an nvidia graphics card. I try to avoid closed hardware as much as possible. My motherboard uses a Via chipset as opposed to an nvidia chipset to avoid having to deal with closed hardware. Wireless cards are harder to avoid. Thankfully, OpenBSD is doing its part in that department.

Closed source is one thing that does not bother me too much. No one has to use it if they do not want. Closed hardware does not make sense. How does it make business sense to make it harder for anyone to use with little effort on the part of the company?

Re:What is wrong with the proprietary driver?

[MasterMnd]

I recently installed Gentoo on a PowerBook with an nvidia card. I can't use nvidia's driver because they haven't released a version for PPC. Fortunately the nv driver is available, so I can use the display, except that I have no 3D support. As a result I can't run beryl (which I was really hoping to try out). Also, even more importantly, I can't do suspend-to-ram. Suspend-to-ram works on the ATI PowerBooks under Linux, but my understanding is that the open source nvidia drivers can't resume the card properly so I'm SOL. Fortunately I was able to get suspend2 working so I can suspend to disk, but it's still a poor substitute. If that hadn't worked then I would've ended-up having to go back to OSX on this laptop.

So, yes, open source drivers ARE important.

Re:What is wrong with the proprietary driver?

[mallan]

nVidia has dropped support for cards older than the GForce4. I have a GForce2 with 64MB and TV tuner that would benefit from this driver.

No, NVIDIA has not dropped support for older cards - they support a driver for legacy cards which continues to be updated for new X and kernel releases. Look for the "Legacy GPU" release:
http://www.nvidia.com/object/unix.html

Cards older than GeForce4 were fixed function pipeline parts - programmable pipeline cards are a different beast, so a split in the drivers is reasonable.

Re:What is wrong with the proprietary driver?

[mysticmike]

apt should probabally never even atempt to update a tainted kernel
so thats a bug in apt

the laptop suspends fine with the NV driver (open source but no 3d)
and i've read articles on how to tweak the driver to suspend ( ie i believe it's possible to make it
work but actually want to spend some time writing code)

i know you can get rid of the logo but it reminds me i'm not booting the wrong kernel,

just imagine having a splash screen for every device in laptop what makes nvidia think they're so special.

also note that the latest windows driver from nvidia don't support laptops so the windows driver
is an outdated version from dell. you can install an up to date windows nvidia driver using an
inf file from

http://www.laptopvideo2go.com/

but trying to program 3d graphics is made more difficult by it being closed hardware
i'm just left guessing wether i'm putting load on the cpu or the gpu, you do stuff
in openGL with no real idea whats done by hardware or software.

if the driver was open source then the installer would give me a working driver like it does for
every other device in the machine.

Re:What is wrong with the proprietary driver?

[picob]

you can reinstall a kernel without black screen with:

NVIDIA-[...].run -a --kernel-name=[NEW KERNEL NR] --no-x-check --no-runlevel-check --kernel-module-only

NEW KERNEL NR is the number for the new kernel as in `uname -r`

Re:What is wrong with the proprietary driver?

[Anonymous Coward]

I grabbed the latest nVidia drivers for my geForce4 ti4200 the other day, and that's been moved to legacy as well. Looks like you need a geForce 5xxx+ series card to use the current non-legacy driver set.

Re:What is wrong with the proprietary driver?

[just_another_sean]

Cool. Thanks for that. Filed away in a README along side the installer...

GIVE US MOHNEYZ

[scenestar]

Only to have the project canceled due to the fascist DMCA .

This better be clean room reverse engineering.

Re:GIVE US MOHNEYZ

[pakar]

Hey, dont think we are all under the rule of the DMCA just becase you are crippled by it :p

Great...

[Kalriath]

If a manufacturer refuses to help the Linux community by providing drivers, wouldn't it make more sense to simply, oh I don't know, boycott their products?

Instead someone has the stupid idea to INCREASE nVidia's market share by getting a community nVidia gives the finger to to buy their products.

Way to encourage companies to support the open source movement... it's basically saying "don't bother writing drivers for Linux, we'll do it at OUR expense!"

Lunacy of epic proportions.

Re:Great...

[Sinryc]

You may think that Nvidia gives its users the middle finger, but if thats the case, ATI is getting the lube ready while pushing you so you bend over.

Re:Great...

[Kalriath]

Hmm, you're right. I was under the impression that there was no 3D drivers for Linux. So, why are they writing OS drivers again? If the proprietry drivers work well enough for Windows users why is that not good enough for the Linux community?

Re:Great...

[Sinryc]

I'm not a big Linux user, actauly, I barely know how to do anything... and thats why I know it isn't ready for the big time(Linux), however, I think most stuff works, just not all for Nvidia, or its not optimized or something. Also, if the drivers are opened up, then any distro will be able to use them no matter what(If they edit them), and if they are opened up, then it means that they can make drivers for other things(Like a new OS thats not linux derived or something)

Re:Great...

[Anonymous Coward]

At the moment there seems to be only negative actions to take (ie a boycott of products without OS drivers). Perhaps a positive example where the presence of a fully implemented open source driver creates a competitive advantage for nVidia will push its rivals to release specs and code for OS drivers for their own products. Maybe it won't work out so optimisticly, but at least this project is making an effort to change the status quo.

Re:Great...

[bcmm]

We can't boycott they're products effectively. All manufacturers of GPUs that one could reasonably use for gaming/AIGLX-type-things are just as bad, and in any case, nVidia is not in a position to open it's drivers. If there were a good open-source supporting GPU manufacturer, we could all go and buy their cards, and nVidia and ATI might start thinking about opening the drivers. The Open Graphics Project is trying to create such a GPU, but it'll probably be a while before they have anything but a rather niche device (they currently have a development card with an FPGA in place of the GPU).

Re:Great...

[dedazo]

I don't play games on Linux (duh), but I think the nVidia drivers I've used so far in RedHat/Fedora are perfectly acceptable. As long as they give me the resolution I want and are stable, I'm jiggy. No worse and no better than any other driver I've used in Windows (not including when actually playing games, of course). I don't think nVidia will ever open up their code, and I think that's a "problem" the GPL crowd will probably have to live with unless they start making their own hardware. I realize there are issues with binary blobs in an OS like Linux, but I think most people can live with them. They seem to, so far.

Re:Great...

[TheRaven64]

We can't boycott they're products effectively. All manufacturers of GPUs that one could reasonably use for gaming/AIGLX-type-things are just as bad Half true. I tend not to play many games, but the latest Intel chips are fine for many AIGLX-based applications. They have pixel shaders in hardware, but not vertex shaders, which is fine for a lot of things, since things like GPU-based image processing all use pixel shaders.

If you're running OS X on a MacBook (Intel GPU) then all of the CoreImage functions (ripple effects, etc) are all done on the GPU, not the CPU.

Re:Great...

[bill_mcgonigle]

If a manufacturer refuses to help the Linux community by providing drivers, wouldn't it make more sense to simply, oh I don't know, boycott their products?

Yeah, but who's the alternative? Last I checked Matrox had checked out of the 3D arena and was still selling old cards for $200.

I suppose a purist might legitimately adopt the "I'm not using 3D until a vendor comes along with OSS drivers" but that also helps cement the OSS display layer as permanently behind (if everybody signs on), whereas now they're running neck-and-neck with Apple.

$10,000

[Swimport]

$10,000 doesnt go far in the development world. I bet Microsoft couldnt make a font for $10,000. Anyway, as we all know, these Linux programmers arent in it for the money. Now if only this gets done and a few other projects, and Ill be running Windows binaries, and playing directx games on Linux, and deleting Windows. Well heres hoping...

Re:$10,000

[Anonymous Coward]

Wine already supports a lot of directx 9. directx 10 is being worked on.

Re:$10,000

[Anonymous Coward]

Microsoft don't make fonts. They buy them from companies like Bitstream and then let suckers like you think they made them.

Re:$10,000

[netsfr]

Funny you say this. A few days ago on the news they were interviewing the cheif sound guy for Windows Vista. He spent something like 2 years to develop the default "beep" for vista. How much did that cost?

Re:$10,000

[Swimport]

Microsoft don't make fonts. They buy them from companies like Bitstream and then let suckers like you think they made them.

Ya they really sucked me on that one. Well at least they didnt buy DOS, or steal Windows. At least I don't pay for Windows.

seems like it's not getting enough mention

[eneville]

Why doesn't this get a more prominent mention on the /., freshmeat and osdn sites? it's for a good OS cause.

They don't need money...

[Anonymous Coward]

http://nouveau.freedesktop.org/wiki/NouveauCompani on_11

[...] The pledge mentioned is however not supported by our project. We currently don't need any money and the person who set it up is not connected to our project.

Congratulations to everyone who pledged to throw money at something that doesn't need any.

Mod parent up

[Kalriath]

The pledge mentioned is however not supported by our project. We currently don't need any money and the person who set it up is not connected to our project. That's awesome. Now if more people browsed with Threshold -1 (or, you know, researched... but that's probably too much to ask!) the guy scamming money wouldn't get so rich.

Save the money, and use it better

[Anonymous Coward]

Wouldn't all the time, effort, energy and money be better spent developing the ATI drivers, which have already made significant progress?

Let's face it, NVidia will always have better closed-source drivers. That's what they do, and that's what gives them the market lead. And for as long as this is the case, I, and the company I work for (around 600 Linux workstations used for graphics) will remain using the closed-source drivers.

Save the $10,000, and when NVidia goes down the shitter, we can put the cash towards buying the *real* code from them.

Better use for the money

[UED]

Rather than spend money on a bunch of RMS lunatics that would sooner die than reverse engineering decent drivers, The money should fund mafia groups to hold Nvidia at gunpoint and force them to release the source code of their drivers. Now that would get opensource drivers with style and since only a few Nvidia guys would be killed there would be no loss of innocent life.

Re:Better use for the money

[Anonymous Coward]

"That's a nice piece of code that you've got there. It'd be a shame if you kept it to yourself..."

And I'm not sure if I support this or not... But it might just be crazy enough to work. I mean, we couldn't use that code, but we'd know how it works and could easily be re-written.

Open Graphics Project

[chill]

Just to bring this back into view...

http://wiki.duskglow.com/tiki-index.php?page=OGPN1 7&PHPSESSID=629ef486f166fab6ef8951de2a5ae96c

The Open Graphics Project is making steady progress.

Re:Open Graphics Project

[MiKom]

The Open Graphics Project aims at 20-30 fps in Quake 3

Re:Open Graphics Project

[x3nos]

Indeed

http://www.theinquirer.net/default.aspx?article=36 761

Re:Open Graphics Project

[chill]

Yes, at 1280x1024 which is better than a Radeon 7000. Don't knock it. It isn't supposed to be the flaming best, just good and open.

Re:Open Graphics Project

[MiKom]

I just wanted to admit that OGP is not an option for gamers. I think it's good project for server market but gamers are left only with nvidia and ATI. However on OGP wiki they say something about increasing performance. If some day I will be able to play modern games with moderate performace, I'll buy it.

Ehh... Have everyone missed the DRM in Vista?

[Anonymous Coward]

There will be NO open source of open specs for future nVidia cards. If they release it it will no longer be possible to playback full res video in Vista. I would think that Nvidia thinks that Vista is a better milking cow then Linux.

Does anybody else not see the huge problems....

[mark-t]

.... with this?

For one thing, if it's not worth $10k to nVidia to open up the source code themselvs, then why should it be that the software shouldn't be worth more than that to develop? And if we are looking at somebody who is doing this largely for philanthropic purposes to accept such a paltry sum, then it is just as probable that this person would have been just as able and willing to develop the same thing for free. Giving this $10,000 to the first person to do it also encourages people to compete with eachother, rather than cooperate with eachother, towards a common goal, which is the very antithesis of open source.

Also, if somebody does manage to do this, there's likely to be an inquiry by nVidia, to ensure that it was not inappropriated from them. Finall, by the time somebody manages to win this prize, because the industry moves so fast, it's bound to be obsolete on the latest hardware.

I appreciate this person's good intentions, but this raises so many red flags with regards to expected problems that I would want to steer as clear from it as possible.

Re:Does anybody else not see the huge problems....

[Kalriath]

He's not giving it to any old project, he's giving it to the project (Nouveau) that he's already selected.

Re:Does anybody else not see the huge problems....

[MoralHazard]

if it's not worth $10k to nVidia to open up the source code themselvs, then why should it be that the software shouldn't be worth more than that to develop?

You misunderstand why NVidia refuses to open their driver code. They're not just being dicks, and they probably aren't too scared to expose their own proprietary technologies, because there ARE benefits to gaining the acceptance of the OSS community that translate directly into more profit.

The real problem is that NVidia didn't write all of the driver code itself, from scratch. They incorporated copyrighted code licensed from other companies, probably some derivative and some word-for-word. They probably use patented algorithms and technologies, too. Several other companies hold rights over the NVidia source code, in such a way that NVidia can't just release it and contribute it to the Linux kernel without violating its own agreements, or exposing the Linux kernel to litigation.

(I really wish I had a quote about this, I cannot for the life of me remember where I read it, but I *believe* that an NVidia exec stated as much in a press conference or release. So unless you do your own research, take me with a grain of salt.)

there's likely to be an inquiry by nVidia, to ensure that it was not inappropriated from them

I find this highly unlikely. Unless one of the companies holding licenses over NVidia's head assumes that they contributed under the table to the OSS effort, there are no grounds whatsover to assume that code was "inappropriated" (sic). NVidia itself has no interest in preventing the release of an OSS driver--if anything, it will help sell me hardware and drive their market share upward as they become the "OSS-friendly" graphics card company. (That's worth a lot of fanboy forum toadying.)

NVidia probably could develop a fresh OSS driver for release into the Linux kernel, via the 'clean-room' approach. It would have to hire developers to do it, provide them with enough specs/code to get the job done but NOT provide them with the licensed code or algorithms, and sit back and watch the payroll fly out the door. This would probably cost them a lot more than $10,000, given the complexity of modern GPUs and the relative rarity of the software skills to do this kind of work. But if someone else wants to do it for free, all they'd have to do is stay the hell away and cross their fingers.

Re:Does anybody else not see the huge problems....

[mark-t]

But then they aren't going to be overly inviting on who they accept to help in the project, because doing so would mean there would be less money for themselves.

Re:Does anybody else not see the huge problems....

[Kalriath]

This group has already said they don't need money. They haven't said they don't WANT it though, so we'll see in time, I guess. But it's just worth knowing that they did not ask for (or expect) to get paid for their project - they had no choice in the matter.

Re:Does anybody else not see the huge problems....

[tajribah]

I think that everybody would be happy if they released not a driver, but specs to their HW. Many people would be happy to write a driver and that wouldn't need a licence from third parties, maybe except for patents, who knows in the insane US patent system.

Re:Does anybody else not see the huge problems....

[Magada]

Stop that meme-pushing. There is no need for Nvidia to open the source to their existing drivers, nor is there any need for them to write new "cleanroom" ones. All Nvidia has to do so that open source drivers for their hardware appear is to release the hardware API. No more, no less. There is no shortage of skilled programmers in the F-OSS world. When asked why they don't do that, hardware companies usually trot out the "trade secrets" line, implying that someone could reverse-engineer, say, a R300 chip if they found out what registers do what. Now, if that is the case, those trade secrets aren't worth much. If it is not, it's pure paranoia on their side.

but will it run

[Anonymous Coward]

...on my USB line?

Plenty is wrong with the proprietary driver

[Sloppy]

Is there some issue with the nVidia 3D driver implementation that would encourage an open-source reverse-engineering effort?

Yes, definitely. It's not portable (you can only use it on the platforms(*) that nVidia has bothered to compile it for). It's not auditable (you can't easily check it for bugs, root exploits, etc). It's not maintainable (if by some miracle you find one of the bugs, you can't fix it).

Those are some pretty serious practical (not merely idealistic OSFOSS) issues. Show me any user to whom none of 1) portability 2) security 3) long-term maintenance, is a major concern, and I'll be left wondering why that user runs Linux at all. That user might as well use MacOS (or maybe even Windows, if they have a lot of legacy apps/games).

(*) Worse, "platform" doesn't even mean just processor types or the name of a kernel. It even means versions of a kernel. Having a driver for Linux 2.6.y doesn't necessarily imply you have a driver for Linux 2.6.z, let alone Linux 2.8 or 2.4. If nVidia ever decides to drop a piece of hardware and stop compiling a certain driver for newer kernels, then users will either have to upgrade hardware (gee, I wonder if nVidia would have an incentive to make people do that) or else use an old kernel. Ouch!

Re:Plenty is wrong with the proprietary driver

[Anonymous Coward]

Show me any user to whom none of 1) portability 2) security 3) long-term maintenance, is a major concern

Like users interested in short-term cost?

Re:Plenty is wrong with the proprietary driver

[cperciva]

Having a driver for Linux 2.6.y doesn't necessarily imply you have a driver for Linux 2.6.z, let alone Linux 2.8 or 2.4.

This sounds like an incredibly good reason to not use Linux.

Re:Plenty is wrong with the proprietary driver

[mandelbr0t]

If nVidia ever decides to drop a piece of hardware and stop compiling a certain driver for newer kernels, then users will either have to upgrade hardware (gee, I wonder if nVidia would have an incentive to make people do that) or else use an old kernel. Ouch!

More appropriate would be to say "or else use a kernel you don't want to." It's just as much of a nightmare being forced to upgrade your kernel as well. Gaming is very sensitive to kernel version (just read the Cedega release notes re: versions 2.6.9 and 2.6.10). Upgrading from 2.6.15 to 2.6.16 caused some Cedega-supported games to stop working.

My major issue with the binary driver is security. Because the driver is a kernel module, remote exploits of the NVIDIA driver will hack the kernel every time. Online gaming brings new life to the idea of remotely exploiting the NVIDIA driver, and not having an auditable driver is a big issue. It took them over 2 years to fix a reported, remotely exploitable issue. It's unacceptable to be forced to use such crap. The only other alternative is to use some other crap which suffers from exactly the same problems. I wish something would shake up the 3D market, but somehow I doubt this project is going to unseat NVIDIA. :(

mandelbr0t

Re:Plenty is wrong with the proprietary driver

[Sloppy]

The only other alternative is to use some other crap which suffers from exactly the same problems.

Or use some other 'crap' that has a different problem (performance).

Re:Plenty is wrong with the proprietary driver

[sjg]

If it is indeed a problem in practice, then it alone is plenty of reason.

$10,000 later

[theendlessnow]

int driver(optr, iptr, context)
nv_obuffer *optr;
nv_ibuffer *iptr;
nv_context *context;
{
    /* Driver goes here, TBD */
}

Here's and idea

[Anonymous Coward]

Why don't some linux geeks get together and build an open source graphics card they can sell at cost to linux users?

Re:Here's and idea

[hasbeard]

I don't know what operating system they do their development on, but here some people working toward an open source graphics card. http://wiki.duskglow.com/tiki-index.php?page=Open- Graphics

Re:Open Graphics Project (Huzzah!)

[forkazoo]

This is always a good related subject to mention. I'm still readily awaiting being able to order one of the FPGA cards to show my support!

Not so much

[Sycraft-fu]

Often technologies will have in the license agreement "You can't release this code." You aren't required to like it, but if you sign the contract (and this stuff involves real, paper, signed contracts) you are required to respect it. nVidia and ATi both license a good deal of things for their drivers (S3TC would be an example). They can't just give the finger to these people and do what they want, they'll get sued and they'll lose because there's a contract in place.

Re:Not so much

[bersl2]

Contracts can be renegotiated, yet this requires time, effort, and money, which are things that no company wants to waste, so I can understand why nVidia, et al. don't do this spontaneously. However, we don't even know what the terms of these contracts are: who the other parties are, what technologies they cover, etc. I don't know whether it's because of non-disclosure clauses or some nuance of contract law that is over my head, but whatever the reason, lots of companies are just downright uncommunicative and uncooperative.

Re:Not so much

[cheesybagel]

So does Intel I expect. Yet they manage to open source their drivers just fine...

When will amd come out with open ati drivers?

[Joe The Dragon]

This whould be even better for linux.

Oh my god - they found a vulnerability

[Ash Vince]

I love the way that the full article even links to the vulnerability advisory and makes such a big deal out of it.

Every single piece of software ever written has bugs. Any that run in a secure area of the OS (like the kernel) but that allow input from unpriveldged processes will also have vulnerabilities (they might allow something the shouldn't). The fact that only one advisory has been found is more of a surprise, especially with all the open source fanboys trying to pick holes in the drivers.

Now ideally no piece of software would allow a direct path to the hardware (like direct rendering) from a security point of view.

But from a performance point of view that bypassing of the secure nature of the unix architecture is essential to allow people to try and run games under linux. Since the games are the only thing tying me to windows I welcome and advancements in this regard.

I would like to see an open source nvidia driver that could match the performance of the closed source driver, but I would be very surprised if this ever happens. If it does, nvidia can just change the hardware so their driver is fastest again.

This is the real problem. The actual hardware the driver has to control is not open, it is a closed proprietary product which you have to sign a nondisclosure agreement just to read the specs of. So whether the driver is open source strikes me as being a moot point.

Why discriminate between hardware and software? Why is one allowed to be a closed system but the other has to be open source in order for us to use it? Why not simply avoid all closed products if you believe that strongly in openness.

Re:Oh my god - they found a vulnerability

[Dachannien]

Every single piece of software ever written has bugs.

I dunno. I've never had a buffer overrun trying to print "Hello world".

Re:Oh my god - they found a vulnerability

[chromatic]

Why discriminate between hardware and software?

One of them is, in comparison, trivial to study, modify, and redistribute in volume.

Re:Oh my god - they found a vulnerability

[Ash Vince]

But if the company selling the product dont want you to study, modify or redistribute either part (hardware or software) what has that got to do with anything?

Re:Oh my god - they found a vulnerability

[chromatic]

Some people believe it's very silly, at best, to create artificial scarcity for things that cost effectively nothing to redistribute -- such as software.

Re:Oh my god - they found a vulnerability

[Ash Vince]

Some people believe it's very silly, at best, to create artificial scarcity for things that cost effectively nothing to redistribute -- such as software

But what if that software could be used by you competitors to reduce their overhead by using part of your code to run thier hardware product? What if your competitors could use your code to produce a product that was 100% compatible but cheaper than yours?

The fact is that there are legitimate business reasons for keeping drivers closed source. Especially if you are a company like nvidia or ati that does not want the general public to know the exact technical specs of the product they are buying. We choose to buy the product, should we not be forced to buy it on the terms of the company doing the selling? Surely the entire capitalist system is based on this?

Re:Oh my god - they found a vulnerability

[chromatic]

We choose to buy the product, should we not be forced to buy it on the terms of the company doing the selling?

That's why I chose not to buy their products. Instead, I bought a laptop from System 76 with Intel graphics and no Windows tax.

Re:Oh my god - they found a vulnerability

[Ash Vince]

That's why I chose not to buy their products.

That is exactly what I was trying to recommend if you dont like their business model.

Dont you just love it when you both argue the same point, at least were not boring anyone else to death with this drivel anymore though now this thread is so many days old.

PS - I actually own lots of nvidia products as I think they are the best of a piss poor lot. I used to prefer 3dfx until they were bought because neither ati or nvidia could manage to product an equivalent product for the same price.

nVidia Linux Drivers support x86-64

[spinfire]

One of the problems is that the drivers are x86 only (although there are old and outdated Itanium drivers). Another issue is obsolete video cards. nVidia could one day stop supporting the TNT or GeForce. What do we do then? If there are no open source drivers, we're SOL on updates. If there are open source drivers, then we can make continued improvements when needed.

You are either misinformed or a liar. The nVidia Linux drivers support x86, x86-64, and IA-64 architectures. This is actually one more architecture than they support on Windows (no IA-64 for Windows systems).

I agree it would be nice to see open source replacements for the nVidia drivers, but please lets not spread or further any FUD about the current closed source drivers. nVidia has done a nice job with the drives. I use them without issue on two different x86-64 machines (one AMD, one Intel).

Re:nVidia Linux Drivers support x86-64

[the_humeister]

You are either misinformed or a liar. The nVidia Linux drivers support x86, x86-64, and IA-64 architectures. This is actually one more architecture than they support on Windows (no IA-64 for Windows systems).

Well, if you really want to be pedantic, they support ia32, x86-64/AMD64/EMT64, and ia64. They don't support ia-16, ia-8, or ia-4.

I agree it would be nice to see open source replacements for the nVidia drivers, but please lets not spread or further any FUD about the current closed source drivers. nVidia has done a nice job with the drives. I use them without issue on two different x86-64 machines (one AMD, one Intel).

They sure have. Except for the security hole that was there for about 2 years.

Re:nVidia Linux Drivers support x86-64

[dinivin]

Except that the security hole that was reported by Rapid7 was only present for two driver versions and was not present for two years, despite the multitude of claims. There was a bug, concerning the RENDER extension, that reared it's head in 2004 and affected numerous drivers, not just nVidia's.

Re:nVidia Linux Drivers support x86-64

[dinivin]

And, in case anyone wants a reference:

http://nvidia.custhelp.com/cgi-bin/nvidia.cfg/php/ enduser/std_adp.php?p_faqid=1971

Now can we please stop with the BS complaints that nVidia allowed a known security hole to exist in their drivers for two years.

Re:nVidia Linux Drivers support x86-64

[smoker2]

And yet the driver necessary for my Geforce 2 card, which is still perfectly capable, is now about 4 generations old. Luckily nVidia still host it on their site, otherwise my Myth box wouldn't work properly. If it was open source, then there wouldn't be the drive to force upgrades just because the drivers weren't available anymore.

Re:nVidia Linux Drivers support x86-64

[StikyPad]

You are either misinformed or a liar.

Classy.

Re:nVidia Linux Drivers support x86-64

[cortana]

4 months is 4 months too long.

Re:nVidia Linux Drivers support x86-64

[Peter La Casse]

You are either misinformed or a liar. The nVidia Linux drivers support x86, x86-64, and IA-64 architectures. This is actually one more architecture than they support on Windows (no IA-64 for Windows systems).

Or simply imprecise. To rephrase your parent poster, "one of the problems is that the drivers support the x86, x86-64, and IA-64 architectures only." People on other architectures are out of luck.

Re:nVidia Linux Drivers support x86-64

[walshy007]

so PowerPC, SuperH, ARM, 68k, sparc, etc aren't important eh? (granted you'd have great difficulty finding hardware with a pci bus on some of those)

I happen to own a fair few bits and pieces with the procs above, it would be handy to have the ability to hack something together and put an old nvidia card in some of the systems, where possible.

I think the parent poster above you when he said x86 meant IA-32 x86-64 IA-64 granted IA-64 is a different beast

Thank you!

[Arker]

When I saw this article I immediately thought: what kind of flaming MORON would pledge money to support drivers for Nvidious instead of open hardware? Not to mention the nueveau folks are on record they didn't ask for and don't need the money anyhow.

Re:Thank you!

[petrus4]

I didn't consider that open hardware article terribly encouraging, myself.

a) It uses a vanilla PCI bus, not PCI Express or even AGP, and is listed at 300MHz. In other words, compared to commercial products, it's going to be SLOW. But I know, I know...who cares if it's technically inferior, right? It's *free.* Good luck with that, guys.

b) They specifically mention that it "isn't a gaming card," and then talk about how they think Quake 3 *might* work with it. If it's a 3D card, if they're not intending it for games, then WTF *are* they intending it for? Given the hollow idealism that is also presumably driving this effort, they're not caring about making it something which large numbers of non-autistics will actually want to use. That would be a lot more useful, since if we were able to get a sufficiently large userbase behind it, other hardware vendors might start seeing producing open hardware as the norm.

c) It is, predictably, GPL or commercial only. No mention is made of the BSD license, which means people using one of the BSDs will not be able to use it with their own license. Once again, people behave as though Linux is the only FOSS operating system in existence.

Re:Thank you!

[k8to]

This is intended to be a card that works well for modern needs, that is it will accelerate typical things a modern display system requires, resulting in rapid display of all kinds of text, as well as visual desktop effects and etc. This _also_ makes it a decent candidate for a "3d" layer like opengl to target. So, by shooting to provide decent acceleration for modern and near-future desktop needs, you also get 3d support.

That said, this is a _developer_ board. It's an FPGA, and designed to provide hardware for parties to actively pursue the development of an open video chip, so that as the design matures, a faster, cheaper, custom ASIC can eventually be produced.

Thus, a cheaper, more powerful design should emerge over time that is very well suited to the needs of the modern graphics display environment, even if it is not the most powerful choice for gaming. I would expect the behavior of this card to be completely respectable for tasks like compiz and aiglx, and probably much more reliable. At least, assuming the project hits its goals.

More open, more reliable, and fast enough is a compelling featureset for some, although perhaps not for you. I had originally hoped to purchase this card initially as an FPGA if only to provide some funding to the project, but the time window for upgrades passed for me, and I invested in intel graphics instead. I may still purchase one in the future, but the expense will not be justified for me personally until excellent X features are working on the device, such as featureful Xv, AIGLX, and possibly some perks with fast text handling and/or compressed video decoding.

Re:Thank you!

[Anonymous Coward]

In other words, compared to commercial products, it's going to be SLOW.

If it has free drivers, DVI and performance in the range of a Radeon 7000 without it's bugs then it's more then good enough for me. I don't play Doom 3 and a Radeon 7000 was enough for Blender and Wings 3D.

What's wrong is, we can't work on it

[Anonymous Coward]

You seem to be missing the point of why FOSS is better, in those areas where we have an itch that needs scratching (but not in all areas, that goes without saying).

Graphics itches us like frickin mad, because it's sexy, because it can always be a bit faster, and because there is always room for more features. And nVidia are dead slow in their driver development, which I assume is because of internal human resource constraints.

Well if there's one thing that FOSS isn't short of it's human resources.

Rock on Nouveau, or whoever takes on this work. It's badly needed.

Oh, and the nVidia binary driver fanboys can just go get lost. Like fanboys everywhere, their arguments fall into the submoron category.

nVidia allegedly can't open-source their drivers for 3rd party legal reasons -- fine. In contrast, we can reverse engineer their hardware and clean-room reimplement their drivers, perfectly legally. Since we can, let's do it. End of story.

I'll bet you DRM is behind the driver secrecy.

[meldroc]

It isn't that NVidia or ATI won't release code. They won't even release hardware specs. There are plenty of kernel and X.org hackers out there who would jump at the chance to write open-source drivers for NVidia and ATI cards. But to do that, you need hardware programming information. You need to know which registers in the cards do what, you need to know what opcodes do what. You need to know what data goes in what registers, or to which addresses, and what data the card sends back, in raw binary. The driver developers don't have this information, and without it, they can't write drivers. NVidia and ATI aren't providing this and won't provide this, citing the need to protect trade secrets. Just to provide the 2-D open-source drivers that X.org does have for NVidia cards, the X.org developers had to run the driver source files through a code mangler that makes those particular .c files look like an entry to the Obfuscated C Contest, or NVidia wouldn't provide enough information to do even 2-D acceleration. The whole point of Nouveau is the laborious process of reverse-engineering NVidia's cards to figure out this information.

Sure, there may be some secret sauce in there that makes for shinier 3-D graphics at a higher frame rate. But I suspect that shiny graphics aren't on the top of the list of things they're protecting. It's DRM. Macrovision's built into every video card that has a TV output port (so you can't use a VCR and tape a DVD movie.) Soon, HDCP will be built into every new graphics card so you can watch HD-DVD and Blu-Ray movies without being able to exercise Fair Use legally. And very likely, all you have to do to turn off Macrovision and completely piss off the MPAA is flip a single bit in a particular register. And it's likely that if hardware programming information was known about newer cards, cracking HDCP would be trivial.

That's why we're stuck with proprietary drivers.

Re:I'll bet you DRM is behind the driver secrecy.

[Anonymous Coward]

But I suspect that shiny graphics aren't on the top of the list of things they're protecting. It's DRM. Macrovision's built into every video card that has a TV output port (so you can't use a VCR and tape a DVD movie.) Soon, HDCP will be built into every new graphics card so you can watch HD-DVD and Blu-Ray movies without being able to exercise Fair Use legally. And very likely, all you have to do to turn off Macrovision and completely piss off the MPAA is flip a single bit in a particular register. And it's likely that if hardware programming information was known about newer cards, cracking HDCP would be trivial.

Isn't it just "ducky" how "clever leveraging" works? Lockdown city, bummer! :( Take down by proxy + bonus: hands appear totally clean.

Re:I'll bet you DRM is behind the driver secrecy.

[Anonymous Coward]

I think DRM is a minor reason for pre-HDCP cards, which is most of them. I think the bigger issue is that neither ATI nor Nvidia want people to see just how many software workarounds they have for hardware bugs...

What's In It For Me?

[Doc Ruby]

Will this project produce a driver that will let my Inspiron 8000 offload window rendering from the CPU to the nVidia GPU, so Ubuntu runs faster?

Or is it just a way to get higher FPS on 3D games running on nVidia HW?

Re:What's In It For Me?

[kiddygrinder]

Unfortunately it won't increase the speed of games; using this driver will probably significantly reduce the speed of games in fact. the main reasoning behind it is that it will provide a stable open source driver that can be distributed with the operating system that at least partially makes use of your hardware, and is better than just defaulting to a 2d driver with no real acceleration at all.

Re:What's In It For Me?

[Doc Ruby]

But will it accelerate my basic desktop by offloading rendering from CPU to GPU?

Re:What's In It For Me?

[kiddygrinder]

Yep, it should do pretty much the binary driver does, though mebee some of the fancier 3d effects might not be implemented

Exactly

[ChunderDownunder]

A similar thing has happened with Java. A few programmers, some employed by RedHat, got together to produce clean-room implementation of the class libraries under the classpath umbrella.

Outsiders scoffed at the insurmountable task they were undertaking, saying it was a waste of time given Sun's implementation.

Now, with nothing to lose, Sun is on the verge of releasing Java under the same license that classpath uses!

The article title is incorrectly capitalized

[Anonymous Coward]

The correct way of writing it is NVIDIA (see NVIDIA website)

Re:The article title is incorrectly capitalized

[Anonymous Coward]

Who cares? You contributed nothing to the discussion.

Goal Reached

[Anonymous Coward]

The site just reported they hit 1000, so yay!
I think this is great, using the binary driver has always been a pain.

My closed source NVidia driver works fine

[jopet]

That is more than can be said for a lot of open source drivers for things like printers or wireless cards. And for many hardware devices no driver exists at all.

I'd rather have closed drivers that work for these devices under Linux than some crappy open source drivers.

What is wrong with that?

Re:My closed source NVidia driver works fine

[frogstar_robot]

I'd rather have closed drivers that work for these devices under Linux than some crappy open source drivers.

What is wrong with that?

1. nVidia can change their minds about Linux support at any time.
2. People may want the hardware to be usable on other arches than i86.
3. It'd be nice to be able to distribute a complete working nVidia Linux system legally.

What is wrong with any of that?

Re:My closed source NVidia driver works fine

[Anonymous Coward]

1. nVidia can change their minds about Linux support at any time.

We'd still have the old drivers, and someone else may want to step in. It's called "competition".

2. People may want the hardware to be usable on other arches than i86.

Isn't that their choice?

3. It'd be nice to be able to distribute a complete working nVidia Linux system legally.

Oh, sorry, I just thought there were other cards out there.

Re:My closed source NVidia driver works fine

[sowth]

Which drivers are you using? I've tried both the "stable" and beta drivers, and both of them crash a lot. In fact, the beta driver seems more stable! ??? I have to turn off all of the extended features and not do anthing "abnormal" like playing videos on the root window (which I would like to do) just so it will be usable. I also think their crappy libs may have broke some of my apps. BBkeys doesn't work properly on that machine now, so I can't use keyboard shortcuts. WTF?

Crappy open source drivers? What planet are you from? All the OSS drivers I have used have been rock solid, while the closed source Nvidia driver sucks. What good is a 3d video card if you can't use 3d and mpeg acceleration without fear of taking your system down?

Re:My closed source NVidia driver works fine

[jopet]

I have never ever experienced a problem with the NVidia drivers. I have had countless problems or severe limitations (when compared with what closed source drivers on Windows can do) with printers, wireless lan, scanners etc. and many devices (e.g. label printers, certain wireless lan cards and many other devices) do not work at all under Linux.

I'd rather have working closed source drivers for all this hardware than nothing or badly working drivers. But the insane policy of making it hard or even forbidding closed drivers will eventually make the situation worse and harm Linux badly.

Do not get me wrong -- of course I would be happy to see open source drivers for every conceivable piece of hardware, but I think some Linux advocates are just totally unrealistic when they think they can force companies to do that instead of thinking about ways how to at least now allow closed source components so that Linux can continue to gain momentum and popularity.

I don't see how this can work.

[sbaker]

nVidia have been very open about the reasons why they can't OpenSource their code - I think we have to take that as a true statement. It's not going to happen - period.

Can we clone their drivers? Maybe - but it could take years to do that - and no sooner we succeed then we'll discover that there have been four generations of new hardware since we started - and the hardware we can support will be so far behind that very few people will want to use it.

You *might* be able to do this for a relatively simple peripheral like a WiFi card - but graphics chips are probably the most complex (and least standardized) single chip device in existance. The driver has to contain a full-up compiler for the OpenGL shader language for chrissakes! (And no, you can't use an existing compiler or translate to some other language because this is a language that supports 4-way parallel arithmetic and has the bizarrest optimisation requirements imaginable!)

This is a massive undertaking. $10,000 doesn't even scratch the surface of the work involved. I seriously doubt that a cash injection of a million dollars would get you a working, useful driver within a couple of years...let alone maintaining it and continually reverse-engineering the next generation of hardware.

Your driver would probably (by necessity) infringe on a bunch of patents too.

Whilst I'd REALLY like the peace of mind of knowing that there is a working, efficient and up-to-date-with-modern-hardware OpenSourced driver out there - it's *so* not going to happen. We need to find clean ways to wall off the nVidia driver so that it can function without being a security loophole and so it can survive kernel changes and such.

HP's involvement in writing the binary-only driver

[embobo]

Wasn't HP the original motivating force behing getting nVidia to write the binary-only drivers? How did they do that? They must have some sway with nVidia. What would they recommend to persuade nVidia to open srouce their drivers? Would $10,000 help nVidia to do so? If yes then give the money to nVidia and not Nouveau.

Re:HP's involvement in writing the binary-only dri

[Millenniumman]

Would $10,000 help nVidia to do so? Well, it's about .00004% of their annual revenue, and they could get it by taking a few dollars off of every employees salary.

I commit to buying an open graphics card

[jonwil]

I hereby commit to buying a graphics card meeting the following specifications:
1.It must be as good or better in all areas (including shader performance) as my current GeForce FX 5700LE.
2.It must be available with PCI Express and/or AGP since if I upgrade my system I will need PCI Express but right now I need AGP.
3.It must have open source drivers for linux that provide good 3D performance (including the abillity to use all features of the card such as programmable shaders)
4.It must provide windows XP drivers (I have no plans to touch vista on this machine) with support for Direct3D and OpenGL which provide performance, functionality and framerates on a given game/3D software that is at least as good as the NVIDIA drivers I have now on my FX7300LE
and 5.It must be available to buy from the manufacturer right now.

If it has a BIOS burned on the card that runs as x86 code, thats perfectly fine as long as its also available in an x86-64 flavor or can otherwise run when the system is run in 64 bit mode.

Does such a card exist?
How much money would be required to get such a card developed from scratch?
How much money would be required to convince ATI or NVIDIA to release the specifications and/or code required such that one of their current cards could meet this criteria? (I know that ATI used to release specs but the cards supported by the open drivers are no longer sold so point 5 isnt met)

For anyone who suggests the open graphics project, that project doesnt meet any of the stated requirements (3D performance is not going to be anywhere near as good as my FX5700LE, last I heard they arent doing AGP or PCI-E, only PCI and if they can come up with a fully functional windows OpenGL ICD without getting in trouble with Microsoft and/or whoever owns OpenGL and the ICD stuff these days, I will be highly surprised)

Re:I commit to buying an open graphics card

[k8to]

What if the system can achieve better performance than your card using PCI?

Re:I commit to buying an open graphics card

[jonwil]

Firstly, all my PCI slots are full with my PCI WiFi networking card and other cards. Secondly, if I upgrade to a new motherboard with, say, a new intel Core CPU (or whatever makes sense to upgrade to from a 3.4GHz P4) I would be buying one with PCI-E and/or AGP and not PCI.

Re:I commit to buying an open graphics card

[Anonymous Coward]

So, in brief, you're committing to buy Intel integrated graphics? :-) It can be part of your system upgrade you're apparently planning anyway.

Re:I commit to buying an open graphics card

[jonwil]

Integrated graphics is right out, I regularly work with very large pieces of data and I cannot use a graphics card that uses system RAM as video RAM.

Let me start you off...

[ratboy666]

10,000 seems like a lot of money. Its not. I expect a driver dev to get $70+ per hour, this pays for 143 hours.

You are not going to get a driver in that amount of time.

But, I will give you clues. The nVidia chip is pretty high on the OpenGL stack. The chip itself handles most OpenGL primitive operations. It just won't do contexts (nor will the ATI). I don't know the underlying protocol to communicate with the chip, but I would guess it is packet based. Registers would prove far too slow. I would imagine that for OpenGL, VGA, video, and mode support you are looking at almost a thousand "registers" or eqivalents.

It may be possible to catch the kernel level packet interfaces -- mode setting and VGA extension should be reversable via emulation. But this won't tell you what any of the commands do. You could try iterating OpenGL and comparing generated packets... but...

Modern chips typically DON'T implement a fixed-function pipeline. So you will have to figure out how OpenGL shader compiler for the chip works (because you have to know the "machine code").

Good luck for a 4 week driver project. The shader compiler itself is almost a C++ compiler which has to be reversed, the communications format and the packet streams. I would give 10 man-years as a first estimate.

Or, you could try to get the vendors to "be nice".

But I won't do it for 10 grand. Sorry.

Re:Let me start you off...

[Schraegstrichpunkt]

You don't have to get the entire driver functioning perfectly. You just have to get something that plays TuxRacer and maybe Quake 3 well enough to make the card more useful than a paperweight.

Change the kernel

[charnov]

Gee, maybe pulling drivers out of the kernel and maintaining a stable ABI might allow manufacturers to release better drivers.

Re:Change the kernel

[rdebath]

Their drivers won't be any better, still crash happy, still dangerous.

BUT, userspace drivers could allow us to kill and restart the driver when it crashes, keep the old slow ABI around so we don't have to bin the working hardware when the maker tells us to and put the driver behind a 'firewall' to prevent it seeing anything that would upset it.

I can see why they don't want to release the binary interface of the card; it's not like an ethernet card ... "put bytes here and they go on the wire" ... the "interface description" would be a complete description of how the card works and why it works that way. The "driver" and the "chipset" are two interwoven parts of the device. The choice as to which side of the AGP a particular function is executed is not fixed and depends on detailed design choices during the co-development of the driver and the chipset.

To try and put it clearly this interface is just like the interface between the different phases of GCC; the C-compiler syntax trees are the "interface" between the phases and could be described but it's not a fixed interface. The programs on the sides of the interface share code so that they know what will be sent and how it should work.

They are stuck here; they cannot release the drivers because of paper contracts; they cannot release the 'interface' because it would require an exact description of the driver to make sense of it; and violate the contract. If they release what they can they're just taking the piss.

IMO there are two choices that allow both sides to be happy
1) Usermode drivers
2) An opensource daughtercard.

A box for a usermode driver would need care. In some ways a process with a usermode driver would have to be very different (eg control of shared memory in OTHER processes, DMA access to memory). The support for restarting the driver would be very interesting.

The simple fact is that there is no hard line between 'software' and 'hardware' and if part of the 'hardware' actually runs on the host system I don't see a problem for open source. AS LONG AS the 'hardware' conforms to a published interface specification. There is IMO no difference between comforming to 'Usermode driver Version 1.1' and 'ATI I/O Port Version 12.43832' both have advantages and disadvantages but both are perfectly reasonable.

The current situation is not reasonable; the kernel interface between the driver and the rest of the kernel is like the GCC example above it's not "Interface version 1.1" it's "The driver module interface for kernel version 2.6.14-ac13" and because these graphic card are designed in part like 'winmodems' (so they can be both cheap and fast) the makers are basically screwed.

Number (2) would be an expensive little card that basically runs a lightweight version of the driver and means that everything on the host side can be open source and/or firmware blob. It in effect turns a 'winmodem' graphics card into a completely selfcontained one. (ie solve the problem by throwing moeny at it)

So where's the usermode video driver project?

Re:Change the kernel

[Builder]

Ya think? Pity the kernel developers don't.

Beyond just allowing for better drivers, this would allow other ISVs who write software that interacts with the kernel to better support Linux and thus grow the Linux ecosystem. But making developers lives easier and more fun was apparently more important.

I've been ranting and railing about the stable API / ABI issue since the new development process was announced. I now have to wait for my distro to stabilise / patch 'their' kernel six ways from Sunday. Even today if you look at the differences between the SLES and RHEL kernels, they are significant, and they are only going to diverge further.

The new kernel development model encourages, nay, demands fragmentation. Welcome to the Unix wars part deux :(

other OS's

[arifirefox]

This is even better news for other OS's such as BSD, Haiku and Syllable. At least linux has the drivers but think about the others

Re:HP's involvement in writing the binary-only dri

[Schraegstrichpunkt]

they could get it by taking a few dollars off of every employees salary.

What an excellent strategy for losing your best employees!

so all you want

[Anonymous Coward]

is the moon on a stick?

Vendors abandon hardware

[bill_mcgonigle]

Is there some issue with the nVidia 3D driver implementation that would encourage an open-source reverse-engineering effort?

Yeah, vendors stop supporting their hardware. When AMD bought ATI they dropped support for the integrated card on my MythTV box which was a year old at the time. That means no TV-out for X.org 7 or higher.

If it were open source they would be free to stop putting resources into it but the community wouldn't be stuck with unusable hardware. And yes, you do have to keep your OS distros up-to-date if they're on the Internet and yes, mere mortals need distros.

I'll buy FPGA OpenGraphics as soon as released.

[shapr]

I'll buy the FPGA OpenGraphics first release card. It's perfect for graphics, FPGA experimenting, signal analysis, etc. I can barely wait!

Making freedom versus placating proprietors.

[jbn-o]

No, because that's one way in which software freedom is made. Sun wouldn't have released their Java software as free software if there were no real pressure on them to compete with increasingly capable free software Java replacements.