NVIDIA's Andy Ritger On Linux Drivers

tykev writes "The Director of Unix Software at NVIDIA talks about Linux drivers, planned features, development cycle, and the open source Nouveau driver. (The interview is in English but all the comments are in Czech.) Quoting: 'NVIDIA's stance is to neither help nor hinder Nouveau. We are committed to supporting Linux through a) an open source 2d "nv" X driver which NVIDIA engineers actively maintain and improve, and b) our fully featured proprietary Linux driver which leverages common code with the other platforms that NVIDIA supports.'"

Comments

Eat my goatse'd penis!

[Anonymous Coward]

Frosty Goatse Piss

Time for a flame fest

[Anonymous Coward]

Binary vs. open drivers. I say we ask Tiger Woods.

Nouveau

[nxsty]

NVIDIA's stance is to neither help nor hinder Nouveau. We are committed to supporting Linux through a) an open source 2d "nv" X driver which NVIDIA engineers actively maintain and improve, and b) our fully featured proprietary Linux driver which leverages common code with the other platforms that NVIDIA supports.

But what will they do when nouveau is complete, and replaces the nv driver? Will they stop commiting to xorg?

Re:Nouveau

[babbling]

Nouveau probably won't ever be "complete" since there's always new cards to add support for and that sort of thing. If Nouveau gets good enough and is the default driver they will probably start contributing to it. They would save money on driver development, and continuing their proprietary driver would be a bit pointless since everyone would be using Nouveau.

Until Nouveau gets good, I imagine they'll keep pushing the proprietary driver, though.

Re:Nouveau

[aabxx]

No open source is ever "complete", for good and for worse :) Even a dead open source project is only "temporary dead"! (whatever that means...)

Re:Nouveau

[Anonymous Coward]

Why would continuing their proprietary driver be ever pointless???

There is no way that an open-source GPU driver can ever achieve same quality (in terms of performance and compatibility at a given timeframe) as an actively developed proprietary driver. GPUs are very complex devices, and drivers make a huge difference on performance. In order to make a fast driver you really need to know how the target GPU works on a very low level. Naturally, NVidia will not disclose this low level stuff about their GPUs to outsiders.

You can compare GPU drivers to compilers. There is no way that open-source compiler (GCC) will ever produce as good code for new Intel's CPUs as Intel's own compiler (ICC) as long as ICC is actively developed.

In addition, the driver must cope with many different applications and tons of different hardware. Even though NVidia started dropping support for older hardware in the recent Linux driver releases, latest drivers support THREE vastly different GPU architectures: NV3x, NV4x/G7x and G8x. That's why NVidia has huge CPU farms dedicated for testing their drivers under different workloads and with a different hardware.

Re:Nouveau

[Andy Dodd]

Yeah. I interpret their "help but not hinder" comment to mean:

"Some of our license agreements prevent us from contributing 3D support to open source drivers, we like Nouveau but can't legally contribute to it."

One wonders what happens if Nouveau becomes the official xorg driver. Perhaps NV will be able to legally contribute to the 2D aspects of the driver. The binary drivers will likely continue to exist as long as they are encumbered by technologies licensed from other vendors which NV has no direct control over (such as S3TC support).

Re:Nouveau

[b1ufox]

The problems as i look at it is loosely like this. Nouveau being a reverse engineered driver misses to harness complex Nvidia specific GPU features.Yes, mod me down if you feel like but that is the truth, a reverse engineered software for a complex GPU is not simple.And now please don't give me example of Samba and et all, i know it. The way hardware evolves in today's hardware industry Nouveau will find it extremely difficult to sustain the pace to match it.

Moreover reverse engineering being a difficult task, it will take time.

Here is the alternative IMHO, AMD open sources ATI drivers. ATI 3D driver gets into mainline kernel. Nvidia will *have* to open source their driver, which can be possibly merged with the nouveau project to make sure it is constantly up to date with the Nvidia's cutting edge technology advances.

But thats just optimistic me. :-)

Re:Nouveau

[neersign]

What I don't understand is why 3d acceleration cannot be added to nv. why is there a completely seperate 3d driver? and what are the reasons that nvidia will help with 2d on nv but will not help with 3d on nouveau? and, what are nvidia's reasons for maintaining a closed source, proprietary driver if they are helping on open source too? it seems to me that if they waste any effort helping with nv, they might as well throw all of their effort in to it. I'm sure their main reason for maintaining a proprietary driver is to protect "secrets" and the reason they won't help with open source 3d drivers is to protect those same "secrets", but I'd really like them to take the "all or nothing" route, and stop teasing - Either make really spectacular proprietary drivers or help the community create 3d drivers that work.

Re:Nouveau

[babbling]

Do you really think Nvidia couldn't renegotiate license agreements, work around patent problems, or at least release a data sheet for their cards?

It's all a matter of how desirable it is for them.

If Nouveau became the official xorg driver and was decent enough that people wouldn't rush to replace it, Nvidia most likely find a way to contribute to it to ensure that Nvidia hardware has a good quality driver on Linux.

Re:Nouveau

[babbling]

It would be pointless for Nvidia to continue developing their proprietary driver once very few people use it because the Free Software one is "good enough" and installed by default. Nvidia doesn't directly make money off drivers, so in such a scenario it would be in their interests to just make sure the Free Software driver is as good as it can be.

What's the point of spending money developing a proprietary driver if everyone just uses the default Free Software one which works reasonably well?

Re:Nouveau

[Anonymous Coward]

...to ensure that Nvidia hardware has a good quality driver on Linux.

Yeah. Just like they did with Vista.

Re:Nouveau

[babbling]

What happened with Vista was not what they intended to happen. They slipped up, but it's not as though they intentionally had crappy Vista drivers.

Re:Nouveau

[Stocktonian]

Just to clarify things the nouveau project is trying to make an open source 2D and 3D alternative not just a 2D driver.

Does it really matter where the drivers come from? As long as the driver has the features I want, the stability I need and is open source I'll be happy. If I can get 2 out of 3 of those then that's pretty good and I'll be supporting efforts to make it a hat trick.

Hopefully the NVIDIA engineers will see the value of a community behind the nouveau project and find a way to contribute directly to that. If they can make contributions to the 3D side of things too then all the better. I think NVIDIA is one of those companies that is trying hard to work with Linux but the ecosystem surrounding it is not very easy for them to fit in with. You know what's funny is that I've never seen an interview that asks how the community can help NVIDIA or ATI. Things such as a stable ABI for loadable modules for instance might mean they can support more architectures.


-----
Want Linux preinstalled on a laptop with a compatibility guarantee? http://www.linuxnotebooks.co.uk/

Re:Nouveau

[jandrese]

As I understand it, while they own all of the IP for the 2d driver, some parts of the 3d support are licensed from other companies and individuals (who hold the patents). The terms of those licenses forbid nVidia from releasing their code (naturally). In order to get a nVidia to open up the 3d support in the driver entirely, the open source community as a whole (or at least some part of it) would have to negotiate with them and convince the companies to open up like nVidia. Given that they make their money off of licensing IP and perhaps selling software (as opposed to selling hardware), this is highly unlikely.

Re:Nouveau

[MBCook]

I disagree. I think it would be pointless to stop developing it. Remember that their drivers are built on a common code base, meaning it shares quite a bit of code with Windows. They are also heavily optimized to a level that free software may never achieve in a reasonable time frame (simply due to lack of documentation and such). Their closed drivers will probably always have superior 3D performance, and that is why people buy Nvidia cards. If you only want 2D, you don't care most of the time. There is still the layer between the kernel and the OS independent stuff that must be worked on, but the have the other work done anyway due to Windows and OS X.

Re:Nouveau

[evilviper]

It would be pointless for Nvidia to continue developing their proprietary driver once very few people use it because the Free Software one is "good enough" and installed by default.

Tell that to ATI. Their r100/200/300 cards were just about fully supported by free drivers for both 2D and 3D.

Aren't you glad they dropped the crappy fglrx drivers, and started developing the open source ones that everyone was using, since, after all, it would be pointless to do otherwise...

Your baseless assertion doesn't make a very good answer to the GP's question of what NVidia is going to do when NV is obsoleted (not what you think they should do).

Re:Nouveau

[Anonymous Coward]

Free software never gets complete, it has stable releases...

Re:Nouveau

[jonwil]

I wonder if it would be possible to produce a driver that didnt use the patented features (and as such presumably didnt support that relies on the patent) but still implemented the rest of the functionality of the GPU. For example, if nvidia has licensed the S3TC patent, they could remove the S3TC code from the "open source" driver.

Or they could go half way by taking all the code that they don't own and all the code that relies on 3rd party patents where the patent holder has forbidden them from releasing source and put those into the "binary blobs" (like they have now) and make the rest of the code (which they own the copyright to and which is not covered by whatever 3rd party patents they have licensed) available (in much the same way at least one vendor of WiFi cards has put only the bare minimum code that has to remain closed for FCC certification into a binary blob and made the rest of the code open source)

Re:Nouveau

[jonwil]

Is Nouveau using the NV code (or the information contained therein) for the 2D part of their work? If they aren't, why not?

Re:Nouveau

[MoxFulder]

There is no way that an open-source GPU driver can ever achieve same quality (in terms of performance and compatibility at a given timeframe) as an actively developed proprietary driver. GPUs are very complex devices, and drivers make a huge difference on performance. In order to make a fast driver you really need to know how the target GPU works on a very low level.

This is the same tired, fallacious argument NVidia has been trotting out for years: "Developing GPU drivers is much too complex. Why don't you Linux kids go run along and play with your toys, and let us write the GPU drivers?"

Well, it's just wrong. I have used many reverse-engineered drivers for complex pieces of hardware, and nearly all of them work as well as or better than the original vendor's drivers. Most recently, I had the pleasure of trying out the Broadcom 43xx wireless driver, which was painstakingly reverse engineered in the face of an INCREDIBLY recalcitrant vendor that won't release a shred of documentation on their devices to open source developers, even under NDA! And, surprise, the card works better under Linux with the reverse-engineered driver than it does with the Windows driver, which seems to lose the signal quite often. The bcm43xx developer Michael Buesch has even got some evidence that certain parts of their code are implemented more efficiently and elegantly than the original driver.

Naturally, NVidia will not disclose this low level stuff about their GPUs to outsiders.

Why is this "natural"? Intel releases documentation on the low level stuff about their GPUs, Realtek goes out of its way to help the Linux community with Ethernet and wireless IC documentation, and Linksys has released the complete code for its Linux-based routers. If a business is based on continual innovation--rather than maintaining an entrenched monopoly--it is entirely possible to be successful and open at the same time.

You can compare GPU drivers to compilers. There is no way that open-source compiler (GCC) will ever produce as good code for new Intel's CPUs as Intel's own compiler (ICC) as long as ICC is actively developed.

Again, this is only true if Intel holds back some of the documentation necessary to make a good compiler. If they publish complete instruction set information, with accurate timing, cache, and pipeline data (which they have done, for the most part), then making a better compiler is "merely" a question of developer resources and talent.

I suspect that, to the extent that GCC code doesn't run quite as fast as ICC, it's because GCC has higher priorities for its code base, especially ensuring support for an incredible breadth of platforms. GCC has essentially become the reference compiler for a lot of embedded development in particular. Even companies like Broadcom that won't lift a finger to help open source out rely on GCC to build the firmware for their own devices.

Re:Nouveau

[jandrese]

I think they have done your first suggestion. It's called the "nv" driver, and it ships with Xorg.

The second suggestion is interesting, and is in principal similar to what they do now. Parts of the driver are still compiled on the host OS, although usually just the kernel interface bits now. Moving more of the code out of the binary blobs really doesn't solve much though, since the binary blobs will still be necessary for the driver to work. Binary blobs are still binary blobs no matter how small you make them.

Re:Nouveau

[notamisfit]

Because nv is obfusticated crap.

Re:Nouveau

[WilliamSChips]

If you wish to see evidence of that, look at nForce. The default nVidia nForce drivers for Linux kinda sucked, so some open-source people developed forcedeth. When nVidia saw forcedeth was already better than their proprietary Linux drivers, they contributed a lot of code to forcedeth, helping it even more.

Re:Nouveau

[WilliamSChips]

The problem with that is that AMD is bound by the same patent crap nVidia is on a lot of their 3D features and neither one can actually open their drivers because it wouldn't be worth the patent minefield to navigate around. But if Nouveau advances far enough to supplant nv, then nVidia will probably send a few engineers to contribute to Nouveau code like they currently do with nv and forcedeth(nForce network driver for Linux)

Re:Nouveau

[jonwil]

The smaller you make the binary blobs, the easier it is for other people to clean-room reverse engineer and clone said binary blobs (as has been done for some of the WiFi cards with binary blobs)

Re:Nouveau

[ppanon]

My understanding is that GCC doesn't perform as well as ICC because ICC uses a number of patented algorithms in its optimizations which GCC cannot yet legally use. Not sure if somebody in Europe might be able to fork a version of GCC that uses those algorithms for use outside the US. The problem of course is that many of the maintainers of GCC work for Cygnus Software (now part of Redhat?) and are based in the US.

Re:Nouveau

[TheRaven64]

It would be nice, since the community could then add their own versions of the missing parts that don't violate patents, or could simply drop support for those extensions. Can't implement S3TC due to patent issues? Well, there are other texture compression algorithms that are in the spec that are of similar quality. Support those, and suggest developers use them instead.

Re:Nouveau

[jdschulteis]

No open source is ever "complete", for good and for worse :) Even a dead open source project is only "temporary dead"! (whatever that means...) It just so happens that your open source project here is only MOSTLY dead. There's a big difference between mostly dead and all dead. Mostly dead is slightly alive.

Re:Nouveau

[somersault]

Seems to me like the point in having a nice simple, stable, 2D only driver would be for use in servers? Our HP servers have 2 Geforce 4 MXs in them. They don't need any spectacular graphics performance, just enough to render some windows competently.

Re:Nouveau

[somersault]

I must point out that there are not 2 GPUs in each server, was kind of pointless of me to say 2 (2 HP servers).

Re:Nouveau

[MoxFulder]

Very interesting! I hadn't heard this before, but found a thread with some brief mentions of patented optimizations on kerneltrap: http://kerneltrap.org/node/486

Do you happen to have a link to any complete stories that talk about these patents?

Re:Nouveau

[Anonymous Coward]

I don't see your point. You talk about ethernet drivers which are ususally low complexity drivers and use this as an argument to support your 3d driver claims.

People have *talked* about developing 3D drivers for modern GPUs for *years* and go figure they don't exist or are even anywhere near existing today. Its a horizontal market and there is a lot of demand for 3D so I don't see the excuse. I'll believe it can be done when someone actually does it.

As far as Intel... lets see Intel sells general purpose CPUs that need to interoperate within a wide range of operating environments. Intel has to release the specs in order for this to be achivable to the mutual benefit of all parties.

NV and AMD are interested only in winning their bloody GPU war and have no interoperability requirements or constraints. Giving away their secrets/work/whatever you want to call it at this level is of no advantage to these companies and I think this is all pretty obvious. Why is this so hard for some to understand?

Re:Nouveau

[mkro]

Yeah. I interpret their "help but not hinder" comment to mean:
"Some of our license agreements prevent us from contributing 3D support to open source drivers, we like Nouveau but can't legally contribute to it."

That is nice. I read:
"They're doing it by the book, so lawyers will be of no long-term use, and we've been around long enough to understand what kind of publicity that would give us. Besides, Nouveau will not be a useable replacement within the next fifteen years anyway."

But maybe I'm just cynical.

Re:Nouveau

[MoxFulder]

I don't see your point. You talk about ethernet drivers which are ususally low complexity drivers and use this as an argument to support your 3d driver claims.

Full-featured 802.11g wireless drivers are pretty darn complex. I don't believe there is anything particularly "exceptional" about 3D drivers, so I see the comparison as perfectly valid. For years we've seen proprietary software makers concede that "open source can do X, but Y is too complicated." I have no reason to believe that 3D drivers are any different.

People have *talked* about developing 3D drivers for modern GPUs for *years* and go figure they don't exist or are even anywhere near existing today. Its a horizontal market and there is a lot of demand for 3D so I don't see the excuse. I'll believe it can be done when someone actually does it.

Well, the Intel GMA drivers already exist, having being released by Intel as fully open-source. And the open source community is (surprise!) working on improving them.

And the Nouveau Project seems to be making quite a lot of progress on NVidia drivers.

As far as Intel... lets see Intel sells general purpose CPUs that need to interoperate within a wide range of operating environments. Intel has to release the specs in order for this to be achivable to the mutual benefit of all parties.

NV and AMD are interested only in winning their bloody GPU war and have no interoperability requirements or constraints. Giving away their secrets/work/whatever you want to call it at this level is of no advantage to these companies and I think this is all pretty obvious. Why is this so hard for some to understand?

I don't think this is quite accurate. As long as it's only NV vs. AMD, they may not have much reason to release anything. However, this creates an opening for a third player, who will start out behind but gain a competitive advantage by releasing open-source driver.

And at this point, Intel is clearly gunning for that spot. They already have an extensive line of integrated graphics, and they'll be making discrete graphics chips soon. If Intel follows through and releases open-source drivers for these new cards, I expect that open-source developers will flock to them in droves and rapidly improve the drivers. This will put more pressure on NV and AMD.

Well, time will tell which of us is right :-)

Re:Nouveau

[Anonymous Coward]

Linksys has released the complete code for its Linux-based routers

I will point out, however, that they only gave out anything at all after threats of legal action were brought against them (GPL). Their response was also to stop making them running linux and now they run a proprietary VxWorks system. So that's not truly a case of open source collaboration.

Re:Nouveau

[voxel]

You're little "complex" broadcom reverse engineering project is but a spec of paint on the moon that orbits our planet earth in terms of complexity when compared to the multi-megabyte nvidia driver binaries.

You silly little man.

Re:Nouveau

[the_greywolf]

Do you really think Nvidia couldn't renegotiate license agreements, work around patent problems, or at least release a data sheet for their cards?

No, I don't think NVIDIA are in the position to renegotiate a license that they don't have direct involvement in. I'll give a few examples from OpenGL extensions for why I think this way:

• ARB_occlusion_query -- HP claims the IP to this extension and licensed it to the ARB.
• ARB_fragment_program, ARB_fragment_program_shadow, ARB_vertex_program -- Microsoft claims ownership of the IP. What, precisely, they claim they own, I don't know. They're not likely to favor opening related source.
• ARB_texture_float, ARB_color_buffer_float, ARB_half_float_pixel -- SGI patent #6,650,327
• ARB_point_parameters, ARB_multisample -- "TBD," according to NVIDIA. It's probably not clear just who has the claim on it.
• EXT_texture_compression_s3tc -- S3 owns the IP on this specific compression scheme. It's the most popular and most widely-used compression format, so removing it might break a lot of games.

A number of other OpenGL extensions are NVIDIA proprietary, and most of the suits will probably hang on with their dying breath, if typical corporate behavior is any indication.

It'd be nice if they opened the driver, but half of the OpenGL ARB's members would probably sue their pants off if they tried.

Re:Nouveau

[UnknownSoldier]

Having implemented OpenGL on the Wii, I think the truth lies someplace in the middle -- it should be doable. While most open source code is crap, *cough* Mesa *cough* the proof is in the pudding -- it works. The biggest challenge is writing clean and abstract code without sacrificing performance.

--
Why C++ still sucks ... because the standard is too lazy to do things right and remove the "long long" garbage, and I've been using it since the early 90's!

Re:Nouveau

[Anonymous Coward]

Err, Noveaue is "making a lot of progress"? They "kind of have 2D" and "very limited 3D support for extremely lucky developers". That is making a lot of progress?

OpenGL extensions include entire programming models. Wireless cards are a far cry from that.

Re:Nouveau

[KDR_11k]

Tell that to ATI. Their r100/200/300 cards were just about fully supported by free drivers for both 2D and 3D.

Yes but how long did it take? I know when I first used Linux with my r200 I couldn't get any 3d capable driver except the FireGL and I never really got 3d to work.

Re:Nouveau

[init100]

Not sure if somebody in Europe might be able to fork a version of GCC that uses those algorithms for use outside the US.

If my understanding of software patents is correct, I don't think that is necessary. Since you don't patent the algorithm itself, but rather the effect it has when it runs, you should be able to include patented algorithms in a source package and just turn them on or off with a configure option. I faitly recall that some projects do this. This is because infringement only occurs when the code runs, not by having the covered algorithm in code.

Re:Nouveau

[Atriqus]

But there would be a point:

-Official Support-
nVidia has no direct control over the Nouveau drivers. As such, they can make no guarantee to their customers that they will even work, regardless of the quality of work by the Nouveau developers. At best, they could say, "Oh you want to actually use our hardware? Well don't talk to us about it, talk to those guys who are under no obligation to actually make it work." And to a lot of businesses, a guarantee that something will work supersedes whether or not it actually does. Spending the extra cash on internally developed drivers to be able to make that claim actually does help them bring in revenue.

Re:Nouveau

[ppanon]

And, um, just how do you think developers in the US will test the code for those patented algorithms that can only legally be used outside of the US?

Re:Nouveau

[init100]

They can't. That they can't do this does not make a fork necessary though.

Re:Nouveau

[Andy Dodd]

Regarding the last one (S3TC) - removing it WILL break a lot of games.

Unreal Tournament's need for the last one marks the beginning of ATI providing proprietary drivers and partial specs for their cards rather than full specs. There was a huge to-do about Unreal Tournament 2003 only working on NVidia cards in Linux back when UT2K3 was released. Epic's response was basically "we use this feature, it's implemented in all Windows drivers and the NVidia 3D drivers, we cannot work around the lack of this feature in the ATI drivers."

Shortly after, the first binary ATI drivers were released, with the main difference between them and the open-source drivers being S3TC. They have diverged since then, as clearly a number of other encumbered extensions have gone mainstream after looking at your list.

Re:Nouveau

[nyu1]

Their closed drivers will probably always have superior 3D performance, and that is why people buy Nvidia cards.

In my case, I buy nvidia cards because intel cards are not available in stores (I'm in Europe), and ati cards have exactly the same problem. I don't care if performance is bad if there's a free driver I can use, because any card without a working free 3D driver is worth as much as a 2D card to me (I'm writing from one of those "2D cards" atm..).

Re:Nouveau

[nyu1]

Things such as a stable ABI for loadable modules for instance might mean they can support more architectures.

They can support all architectures with a free driver that can be recompiled on ABI change and easily adapted on API change. Providing a stable ABI would only archieve reducing the incentive they have to release that free driver. Even worse, it would mean the community accepts the binary blob, and make liberation of that code much more difficult in the future.

see no evil, hear no evil, do no evil

[Anonymous Coward]

So, basically, it's "we don't care, we're not changing". It's kind of sad that they refuse to release open drivers, although not nearly as bad as ATI's fglrx, the nvidia driver has issues too and seems to be used to force upgrades (semi-recent drop of support for 4xxx cards, eh?), basically meaning that otherwise fine cards will be unusable in a couple months or after a few kernel release cycles.

Re:see no evil, hear no evil, do no evil

[WilliamSChips]

Even if they want to, they can't. Or rather, they don't want to enough to navigate through the patent minefield laid out by companies such as SGI.

Move along, nothing to see here

[dfdashh]

Still no 3D support without having to use a proprietary (closed) driver. Wake me up when that changes.

Re:Move along, nothing to see here

[Anonymous Coward]

Still no 3D support without having to use a proprietary (closed) driver. Wake me up when that changes.

Sure thing, Mr. Rip Van Winckle!

Re:Move along, nothing to see here

[aabxx]

Will an attitude like that really help anyone?

There are understandable reasons for nvidia wanting to keep it like it is.

Re:Move along, nothing to see here

[Anonymous Coward]

you can never get 3D in a computer monitor, they have height and width there is no depth, although some fancy graphics can simulate that third dimension however it is not real, i say let Nvidia develop 2D to the best of its abilities and let the game developers continue to fool idiots in to thinking there is a third dimension on their 2D monitors...

Re:Move along, nothing to see here

[Ngarrang]

Still no 3D support without having to use a proprietary (closed) driver. Wake me up when that changes. What is wrong with a closed-source driver? We are not talking about a business application or web browser client.

Re:Move along, nothing to see here

[Anonymous Coward]

"What is wrong with a closed source driver?"

Don't you know RMS's printer driver story?

Re:Move along, nothing to see here

[tomservo84]

Two words: orphaned hardware.

When nVidia decides they no longer wish to support an older piece of hardware, they just "build it out" of the driver. The driver no longer works with your perfectly good video card.

If the driver was open sourced in *ANY WAY* (including the scheme from somewhere above where they license it only for use with nVidia hardware) the code would aready be out there in the open and would always be accessible.

Re:Move along, nothing to see here

[Vanders]

A binary Linux driver doesn't help much if you're not running Linux.

Re:Move along, nothing to see here

[neersign]

the problem is that they can't even release a properly working closed source driver. And since it is closed source, all that we can do is sit back and hope that in the next release they might address the issues we are having, if we are lucky. If the driver was open source, then anyone would be able to fix the problem, and submit the change so things would get fixed quickly. If the nv driver and the nouveau driver show anything, it is that people are willing to expend their own time to develop a driver that fits their needs, so if the nvidia driver was opened, there would definitely be people willing to help develop it for free.

Re:Move along, nothing to see here

[Hal_Porter]

you can never get 3D in a computer monitor, they have height and width there is no depth, although some fancy graphics can simulate that third dimension however it is not real, i say let Nvidia develop 2D to the best of its abilities and let the game developers continue to fool idiots in to thinking there is a third dimension on their 2D monitors...

Yeah, I don't like it when artists use evil tricks like perspective too! They should be honest like they were back in ancient egypt.

Re:Move along, nothing to see here

[Anonymous Coward]

What is wrong with a closed-source driver? We are not talking about a business application or web browser client.

Having the source means you can compile the driver for different architectures. Go ask PPC Linux users how important that is.

With the source, you are no longer at the mercy of a company releasing another binary blob when they can be bothered, every time the kernel changes. Currently, your nvidia driver will break after upgrading the kernel.

Having multiple experts going over the code will probably improve it, giving everyone a better product.

nvidia makes money selling hardware, what they spend on software hiding bugs in the hardware and cheating performance tests (as well as ATI), is money down the drain.

Remember, linux users want a stable system. They don't get their willies out and start masturbating over a couple of extra fps on the latest energy guzzling video card.

Re:Move along, nothing to see here

[drsquare]

What's forcing you to use the new updated drivers?

Re:Move along, nothing to see here

[sparcnut]

Yeah, like our eyes can truly see 3d anyway... computer displays can take full advantage of the capabilities of our eyes wrt. 3d vision using stereo glasses.

Re:Move along, nothing to see here

[TheRaven64]

How about the bug in the nVidia drivers last year that allowed arbitrary code execution by anyone who can make you display an image? I wouldn't run code like that, especially not with kernel privilege. The fix? Update to the latest drivers. Oh, but they don't support all of the older cards...

I guess it's a choice of disable 3D support, keep a remote vulnerability, or buy new hardware.

Re:Move along, nothing to see here

[WilliamSChips]

Well, you see, I need to use both the old driver for my old hardware and the new driver for my new hardware, but there's no overlap in kernel versions!

Re:Move along, nothing to see here

[Anonymous Coward]

That has got to be the stupidest, most ridiculously idiotic AC Troll I've ever seen in my life.

Protip: It says "Anonymous Coward", not "Anonymous Fucktard".

Did they mention 'why'

[QuantumG]

Every interview with the NVIDIA "developers" they repeat the company line (see summary) but fail to state WHY this is the company line. Speculation abounds as to why this strategy was chosen, but I've yet to see an official comment from NVIDIA.

Anyone?

Re:Did they mention 'why'

[Kjella]

Licensed code and patented algorithms that require closed-source implementation, I'd wager. Or just a general paranoia and protection of software implementations. Or that they're still the preferred choice over ATI for Linux machines. If we ever get to the point where you say "Go with ATI instead, their drivers are much better *and* open source" maybe they'll change their opinion. Until then, they're probably committing most of their resources to beating ATI in the Windows market. Can't say I blame them for that.

Re:Did they mention 'why'

[QuantumG]

More speculation, just what we need. Hopefully the next guy who interviews an NVIDIA dude will ask the question and we'll get something other than bullshit back.

Re:Did they mention 'why'

[ajs318]

Possibly the fact that changing one bit in one byte is enough to turn a £30 graphics card into a £300 graphics card. Or maybe that they ripped someone else's code off.

Re:Did they mention 'why'

[simm1701]

Patents wouldn't exclude an open source release. Patents they want to enforce would preclude using the GPL or BSD licenses for example, but it would certainly be possible for them to release their driver open source with a license they are happy with that protects patents they own.

Its the licensed code, as you mentioned, that causes the problem.

They would need permission and an agreement on license type from every single company or organisation they have licensed code from

Re:Did they mention 'why'

[QuantumG]

Or maybe the secret of excelsior is in the code and they don't want to dethrone the new king. Speculation == pointless.

Re:Did they mention 'why'

[iainl]

No, that's the ATI drivers where the cheap cards are clock-locked in software for Windows...

Although I suppose NVidia could be doing it as well.

Re:Did they mention 'why'

[davFr]

AT least the "third-party closed-source code" excuse is BS.
if one looks at the last version of nv driver before it got obfuscated (upon nVidia request), you will see that the driver code was actually doing close to nothing:
http://cvsweb.xfree86.org/cvsweb/xc/programs/Xserv er/hw/xfree86/vga256/drivers/nv/Attic/nv3driver.c. diff?r1=1.1.2.5&r2=1.1.2.6&hideattic=0&only_with_t ag=xf-3_3_3

Re:Did they mention 'why'

[orzetto]

Every interview with the NVIDIA "developers" they repeat the company line (see summary) but fail to state WHY this is the company line.

NVidia is headquartered in Santa Clara, CA, USA. This means they are under a software patent legislation. They are either in breach of patents and they know it, or they are and they do not know it, but suspect they are. They have a competitor, ATI, which has likely a bunch of patents they are infringing (and they are probably doing the same to NVidia). Opening the source would be inviting ATI to dispose of its main competitor by lawsuit.

NVidia could relocate to a software-patent free legislation (Europe? South America?), but expecting them to move a whole company of over 3,000 people for open source is kind of utopistic. NVidia is not Canonical. Also, they are a hardware company, and just publishing their specs (not the driver source, just the card specs) may expose them to patent lawsuits if some spec reveals certain implementation details (and this time there is no free zone from hardware patents).

So, until a small upstart from Germany or Mexico fills the niche by selling 3D cards with available specs or maybe even FOSS drivers, we are pretty much stuck with this situation. Alternatively, wait for Nouveau, but its development is based on reverse engineering and will be inherently inefficient compared to simply implementing a spec. Thankfully it is just 3D acceleration, think of the mess if we had this situation in something more important like ethernet interfaces.

Re:Did they mention 'why'

[Anonymous Coward]

If we ever get to the point where you say "Go with ATI instead, their drivers are much better *and* open source"

You mean like people currently say "Go with Intel instead, their drivers are much better *and* open source"? :)

Re:Did they mention 'why'

[QuantumG]

Yes. The key point of my post was that people like you are speculating and this is completely useless for figuring out how to convince NVIDIA to open source their drivers. We need a statement from NVIDIA on why they choose not to open source their drivers so we can address that statement and work with them to do the right thing here. Addressing stuff they didn't say is just pointless.

Re:Did they mention 'why'

[Chandon Seldon]

Wait a second... Nvidia requested that the "nv" driver be intentionally obfuscated and the X.org people agreed? That's absurd.

Re:Did they mention 'why'

[nyu1]

So, until a small upstart from Germany or Mexico fills the niche by selling 3D cards with available specs or maybe even FOSS drivers, we are pretty much stuck with this situation.

intel already does this.

black window bug

[xNstAble]

Improving our GLX_EXT_texture_from_pixmap implementation's handling of out-of-videomemory scenarios (this is the cause of the "black window bug" when running Compiz or Beryl composite managers) Good to see that this issue is still wellknown to nvidia developers, this bug has been around for a lot of months and has never been addressed yet. This is the typical problem that open sourcing the drivers would solve in a much shorter timeframe I guess..

Re:black window bug

[LWATCDR]

Why do you guess that?
They know it exists but my guess they are having problems duplicating it. Even if it was totally open sourced I doubt that the fix would come much sooner. Frankly until they can say "If I do X,Y, and Z I get the error" it will be very hard to track down. Frankly with the complexity of the hardware they are dealing with it would take years for opensource developers to get a grip on it.
I like FOSS but it isn't magic. Take a look at just about any FOSS project and you will see that a few people do most of the work. Take the NV driver for example. It is FOSS and it still has bugs. Also take note that the NV driver is mostly maintained by NVidia.
The simple reason that NVidia probably doesn't release their drivers as FOSS is because ATIs drivers tend to be inferior to Nvidia's and that gives NVidia an advantage. The why don't you release the specs argument is probably also usless. I have a sneaking feeling that a lot of what makes the NVidia cards work so well is in the driver. Without that information I would bet that the FOSS drivers would suck.
GPL prevents people from cutting and pasting your code. It doesn't prevent a programmer from learning from your code and taking the basic ideas in the code and producing a Closed Source product from the. IT never can unless FSF starts patenting their code and that would just be wrong.
Of course there is another reason that NVidia and ATI might not want to release their code. With all the stupid Software Patents running around I would bet good money that both of them infringe on somebody's stupid patent. Why risk spending a few million dollars on court costs when you don't half too.
I keep hoping that ATI will release the driver as FOSS but I can see many none evil reason why they don't.

Re:black window bug

[3p1ph4ny]

It's not very hard to duplicate (for me at least).

When I'm running beryl 0.2.0 and another 3D graphics intensive application (say, Warcraft III) and I open up *any* window, it's black (although it's decorated). Closing war3 and reopening the window yields a window that isn't black. All windows before war3 is started are not black.

Re:black window bug

[LWATCDR]

Well then they shouldn't have a hard time fixing it.
Are you using XGL or running Beryl directly?

Re:black window bug

[Jussi K. Kojootti]

Take a look at just about any FOSS project and you will see that a few people do most of the work. Take the NV driver for example. It is FOSS and it still has bugs. Also take note that the NV driver is mostly maintained by NVidia.

Let me guess: you haven't looked at nv code or talked to anyone who has? It looks deliberately obfuscated and crippled -- let me quote one of the noveau project goals: "De-obfuscate the nv source". That should tell you why nv is still crap.

Re:black window bug

[LWATCDR]

No I haven't but I have looked at lots of other peoples code. Guess what unless you are using identical styles then your code may look like crap to them. I know one guy that wants every member variable to start with m_ which drives me nuts. If a variable is local to a method or context within a method then it will be defined in close to where it is used. I tend to assume a variable is member variable.
Almost any none trivial piece of code will take a long time to wrap your brain around. Maybe you should take a look at the Intel drivers and see if they are actually any better?

Re:black window bug

[Jussi K. Kojootti]

There is a difference between complex and clearly obfuscated (and I did not use the term with exaggeration in mind), but you don't need to take it from me: Google will show that a lot of xorg/xfree86 developers agree.

I'll add another quote. This was added to the xfree86 sources after the initial nvidia upload:

After we already finalized XFree86-3.3.3 NVIDIA forced The XFree86 Project to replace the sources we had with sources that were partly run through the C preprocessor in order to remove some of the names that NVIDIA thought might give away IP from NVIDIA. This resulted in unreadable and unmaintainable code. The XFree86 Project is strongly opposed to such obfuscated code. We do not regard this as free software according to our standards. Due to the extremely late date of this decision from NVIDIA we decided to include the code as offered by NVIDIA. We are considering to remove support for the later NVIDIA chips in a future release, though.

Re:black window bug

[xNstAble]

Ok, many good points. But also, I can reproduce this bug very easily (indeed is it enough to enable TFP and open a few big windows on my hardware. You can keep maximizing browser windows for a while and tha all new windows get black), and I would be probably likely to be trying to investigate the bug and provide a patch (as I generally do for other bugs annoying me this way.. and also I'm sure many other around could do the same even better if they had the code).
Anyway, agree with you that sw could make the difference between a card and another when talking about the performances of 3d HW.

A question

[Fross]

Why do you _expect_ a company to do business according to your own rules?

Sure, you can vote with your wallet, but it seems to me demanding that a company release their drivers in open source (and let's not forget, that probably contains a vast amount of work they spent their own R&D budget on) is expecting a huge amount of work from them, in return for a small amount of convenience for you.

I'm just wondering why so many open source users have disdain for companies not open sourcing their software when it is potentially against their aims to do so. I mean, to the exclusion of actually using the software which could make their computer experience better. Surely we haven't got that many mini-RMSes?

Re:A question

[Whiney Mac Fanboy]

seems to me demanding that a company release their drivers in open source

I read & re-read the parent comment, but couldn't see them demanding anything. WTF are you talking about?

I mean, to the exclusion of actually using the software which could make their computer experience better.

Some people have quite pragmatic reasons for preferring open software - particularly kernel software. Driver crashes were one of the things that made windows (particularly in the late 90's / early 2000s) such a mess.

If you're making a hardware purchasing decision and want to run linux, of course you should try to buy from a company that supports FOSS.

Surely we haven't got that many mini-RMSes?

Finish the troll with a flourish. Nice work.

Re:A question

[dfdashh]

What part of my comment demanded anything from Nvidia? They have every right to do exactly what they are doing, and I respect that. They have excellent support with their proprietary drivers, and I use them whenever I can. I expected TFA to have some trinket of _news_. I was sadly disappointed, that's all.

Re:A question

[Aladrin]

Because it works! There's a saying, 'The squeaky wheel gets the grease.' If Linux users are very vocal about how they expect companies to do business, those companies -will- feel pressured to move more towards that way of doing business.

It's the exact same reason that lobbying a congressman (without money) works. Once they hear it enough, they know it's important to the people that are most important to them: Their customers. (Or voters, as the case may be.)

When people don't tell a company how to behave, you end up with companies like Walmart. Walmart used to be about the country, the consumer, and the profit, in that order. They gave up on the whole 'made in the USA' thing quite a while back. They gave up on customer service even longer ago. They only care about the profit now. They do it by having cheap goods and cheap wages. For people who only care that the goods are cheap, it's a great store. For the rest of us it sucks.

nVidia has the choice of only catering to the mainstream Windows-based gamers, or also adding on a rabidly-loyal group of fanatics that are willing to work for free to make their business better. All nVidia has to do is LISTEN TO THEM and release their drivers open-source.

Yes, there was a great amount of R&D involved in their drivers, but most of the stuff that makes their drivers 'great' on windows just doesn't apply to Linux, like that massive control panel. That doesn't even exist in the binary Linux driver.

The code doesn't have to be GPL or any such. They could release it under their own license that specifically states the code can only be used for a driver for nVidia cards. The only thing necessary is the ability to improve the code at will. (I think they would find it advantageous to go to GPL later, but that's another discussion.)

nVidia really has little reason not to open their source code to the public, unless they are doing something illegal or extremely unethical in their drivers. (Cheating at benchmarks, etc.)

Re:A question

[mrcgran]

One big problem in not releasing the source code is that they actually are not making our computer experience better: their drivers have bugs, and we will be locked to whatever features and bugs THEY want to make available to us -- so, basically we become hostages of their will, they can do whatever they want, because whithout THEIR driver, your nvidia card isn't worth its weight.

In the future, when new versions and extensions of OpenGL are released, we won't have any guarantee that they will properly update the drivers. So, you'll probably won't be able to use their proprietary drivers in 5 years for new applications (shining new wobbling effects), because these apps will need new extensions, but the driver for your specific nvidia card is arbitrarily not supported anymore by them (they want to force you to throw the old one away). Too bad for you.

On the other side, if we have access to the source code (or at least the hardware specification), we don't even need nvidia's help: we can do the updates/bug-squatting ourselves, much better than a small team at nvidia. This is something that these companies don't get: the whole world is willing to write their drivers for free and maintain them to the end of times, but they refuse the consumer this right (or maybe they get, they just want you to throw away your old card and buy a new one). We don't want a huge amount of work from them, quite the contrary! It's *way* cheaper for them to release an open-source driver: it costs nada/zero, we can build one with the bare bones of a reasonable hardware specification, a little pdf file -- how much does it cost to post a pdf file on the Internet?

There's no RMS ideology in that, only the absolute minimum someone would expect in terms of support for something you bought. Nowadays, the choice is clear: go Intel X3000/X3500, which supports open source and you can be sure will always be up-to-date. Ignore nvidia and ati, until at least one day (I hope so) nouveau arrives.

Re:A question

[RichardDeVries]

Why do you _expect_ a company to do business according to your own rules?

I can only speak for myself, of course, and I don't expect them to do anything. I just don't understand them.
You're saying that it will cost NVidia (and others) money to open source their drivers. It will save them money. Their drivers would be developed, tested and used by more people. Development would speed up and it would cost less.
The fact that there are a lot of people out there who want nothing but OSS on their machines is only one issue. I would _prefer_ an OSS driver, but I _want_ a working driver that's at least up to par with the Windows version. I've got an ATI card so I'm fscked. Open sourcing drivers would give the Linux community state-of-the-art drivers within a few months without extra costs for the manufacturer.
The reason they don't do it, of course, is that they're afraid the design secrets of the hardware will be revealed. I can't really judge if that would be the case (I'm no hardware expert), but I'm guessing that risk would prove minimal. If company A _really_ wants to know how company B's hardware works, they'll figure it out even without the source code of the drivers.
NVidia's case is even more confusing as they do release their 2d drivers. As I said, I just don't get it. Or they just don't get it. If NVidia takes the first step, they'll be embraced by most of the Linux community and that is a fast growing market.

Re:A question

[Ash Vince]

Finish the troll with a flourish. Nice work.

I saw nothng in the parent post suggesting the poster was a troll. You post sounded more like one.

On the subject of RMS though, does he actually contribute any code to any projects or do anything usefull or is he just a worthless commie who whines about other poeple not having the same political outlook as him using Linux.

(That was a troll!)

Re:A question

[Synn]

I'm just wondering why so many open source users have disdain for companies not open sourcing their software when it is potentially against their aims to do so.

Yeah, but it's really not against their interest for hardware vendors. ATI and NVidia sell hardware, not software, so it makes little sense for them to keep their drivers closed source. They're not going to expose any IP they couldn't patent or the competition doesn't already know about.

And the reason Linux users want open source drivers is because they work better than closed source ones. I don't buy ATI specifically because their close source drivers suck. I buy NVidia because despite their drivers being closed source, they do a good enough job with them. If ATI got their act together and put out a good open source driver I'd switch to them in a heartbeat as I'm sure many other Linux users would.

Re:A question

[Yfrwlf]

You've been on Slashdot how long? I'd just think you would have found that out by now. I don't think anyone is "expecting" Nvidia to open source their 3D drivers anytime soon, but doing so gives control back to us, the consumers. Or at least, it helps. :) I know some of you may be used to buying things that are completely controlled by contracts/EULAs/DRM/binaries, but back in the old days, consumers actually had rights to do what they pleased with their stuff, for the most part that is, including tinkering, modifying, selling, duplicating, sharing, etc. :)
 
Equivalent to the open source movement in the physical realm is the Maker Fair http://makerfaire.com/ which promotes basically the same agenda. OK, so the Maker Fair is much much smaller, but still.
 
Just because some of you believe that we should all bend over for the Man because you're used to it, give up our freedoms for security or because some company would LIKE to control most every aspect of your life and force you to watch their dumb advertisements no matter how many times you click the menu button on your DVD remote (yes, it pisses me off), doesn't mean that we should, and that everyone who doesn't like it should shut the hell up about it and accept oppression.

Re:A question

[Fross]

You're saying that it will cost NVidia (and others) money to open source their drivers.

Not exactly - that it cost them money to develop the drivers, including their own R&D into how to best squeeze performance out of their cards, and gain an edge over their competitors. This is highly sensitive information and undoubtedly cost them a lot to develop. To open source their drivers would be to give this all away to their competitors, who you can be sure would be the first to look at the code.

As good as the open source environment is, they do not have the specialists or the resources to develop a cutting-edge product which is, after all, exactly what the video card market is.

The fact that there are a lot of people out there who want nothing but OSS on their machines is only one issue. I would _prefer_ an OSS driver, but I _want_ a working driver that's at least up to par with the Windows version.

Which is the boat I'm in. I'm thankful they spent the effort to make (closed) Linux drivers at all, but I'm finding it hard understanding seriously gimping your system rather than run a closed-source driver, which was the attitude of the original post I responded to.

Re:A question

[Phisbut]

Because it works! There's a saying, 'The squeaky wheel gets the grease.'

I played way too much Baldur's Gate in the past... I've read that and thought "that doesn't sound right..." I just can't imagine that saying to be anything other than 'The squeaky wheel gets the kick'. (As well as 'The bigger they are, the harder I hit').

Re:A question

[Carewolf]

nVidia really has little reason not to open their source code to the public, unless they are doing something illegal or extremely unethical in their drivers. (Cheating at benchmarks, etc.)

All modern windows video drivers "cheat" in games and benchmark. The drivers fixes bugs in the games, and introduces new ones. Since the driver recognizes the game running and adapts accordingly, it would be odd to not do the same with benchmarks.

Re:A question

[Fross]

This is exactly the attitude I'm talking about - the notion that anything you don't like is "oppressing" you, keeping you down. Yeah, stick it to the man!

So you don't like the license a piece of software is released under? You think you have the right to it under a license of your choice. You don't like the DVD with ads in it. So perhaps you should "tinker, duplicate, share" it, as you put it, by removing them and spreading the DVD without them? What about the boring bit in the middle of the movie, perhaps remove that to? After all, it's oppressing you, and you're not going to be bound by the contract. You're coming across like some kid who goes into McDonalds and wants the Happy Meal, but insists on a quarter pounder instead of nuggets, an icecream instead of fries, and a large milkshake, because that is the way you want it, so they MUST give it to you.

No-one is stopping you having an open source 3d driver for your nVidia card. You can write one yourself. But you're telling some company that they have to sell you their product on your terms. And then probably support your driver after you've hacked it to bits. By all means, vote with your wallet, if they don't cater to your tastes, don't buy their product. I hate DVD adverts as much as the next guy, but I understand that it's their business model, and I can either buy it or not. To tell them to not do it because I don't like it is tantamount to telling a musician they HAVE to sell me their new album without tracks 2, 3 and 5 because I don't like them.

I'd like to see you become a creator of content that your livelihood (and that of your family) depends on, then accept someone trying to rip up the way you want to sell it, because it doesn't suit them. You have *choice*, and you have the ability to do your own thing. Your demands others behave the way you want is spectacularily similar to the worst company attitudes about how they want their consumers to behave.

As for how long I've been on Slashdot? If you can't work that out, you MUST be new here.

Re:A question

[pkphilip]

Though one may argue that having the source for a driver may result in the driver becoming more stable over time as a lot of people contribute changes/fixes to it, I feel that this may be overstated.

Consider Nvidia/ATI drivers on Windows or Mac OSX - these binary-only drivers are feature rich (are they more feature rich than their binary-only drivers on Linux?) and most users are quite happy. Bugs do occasionally show up, but they are normally fixed by Nvidia/ATI within a reasonable time frame.

However, I have noticed that these same manufacturers take forever to fix bugs which show up only on Linux.

That indicates to me that the reason that these binary drivers are not that stable on Linux is not because of the binary nature of these drivers but because the Linux user community matters less to Nvidia/ATI than the Windows user community.

And that is understandable - the number of windows users is roughly 93% http://http//en.wikipedia.org/wiki/Comparison_of_W indows_and_Linux/ and Linux users probably account for about 3% to 4% for all computer users.

So it is understandable that a hardware manufacturer prioritizes bug fixes for their larger user base (windows) rather than for the Linux users.

Unless Linux gains the kind of market share which will force hardware manufacturers to take it seriously, we can expect less than stellar drivers and support from them.

Re:A question

[iminplaya]

Why do you _expect_ a company to do business according to your own rules?

It's called bargaining, or haggling. Very common practice in almost all business transactions. There's nothing wrong with it at all. The hardware market is very closed and full of collusion, just like the pharmaceuticals. Without a suitable alternative the pressure must be put on these people to respond to market forces. That's pretty difficult to do when they control the market. And it's the buyer that's supposed to control the market, not the merchant. Real competition would bring both nVidia and ATI around real quick, and you would see completely open drivers. But there is none, so I'm for whatever it takes to force their hand. The buyers must assert their authority and not allow the vendor to dictate conditions.

Re:A question

[AP2k]

that probably contains a vast amount of work they spent their own R&D budget on Fair point. Although it is moot because they arent making a single dime off it because they give it for free.

Re:A question

[Aladrin]

Oddly enough, for just about every saying like that, there's an opposite one. The old-time opposite is "The nail that sticks out gets hammered down." I like to use them because they mean just about nothing, since they all have a saying that negates them, yet most people take it as fact that both sayings are correct, simply because they're said often.

And, er... 'Go for the eyes, Boo!'

Re:A question

[Enderandrew]

ATI is opening up their drivers.

Intel has already open-sourced their drivers.

Other hardware manufacturers have released data specs, and opened up drivers.

What secrets is NVidia really trying to protect?

The lack of solid, stable 3D drives in Linux is a major stumbling point for Linux gaming, and the main reason I dual-boot.

I've been buying NVidia cards because I prefer their Linux drivers, but the second AMD/ATI releases an open-source 3D drive, my 7600GT is going in the trash, and I'm buying ATI cards.

Re:A question

[Enderandrew]

Yes and no. If they open up the drivers, some of that R&D cost is lessened by community contributions. For instance, the Noveau project could probably write the driver fully if NVidia opened up their current binary driver and gave them full data specs.

Then NVidia no longer spends any money on Linux driver development.

The Linux community has a free-standing offer to write Linux drivers for free lest you forget.

And as far as protecting secrets, both Intel and ATI are fully open-sourcing their Linux drivers.

Re:A question

[Anonymous Coward]

How are those sayings opposite? Both refer to some sub-optimal situation bringing attention to itself, and being fixed.

Re:A question

[shish]

nVidia really has little reason not to open their source code to the public, unless they are doing something illegal or extremely unethical in their drivers.

Which category does "not holding copyright for all the code" come under? (I'm suggesting that they've licenced third party stuff, not stolen it)

Re:A question

[mhall119]

You're saying that it will cost NVidia (and others) money to open source their drivers.

Not exactly - that it cost them money to develop the drivers, including their own R&D into how to best squeeze performance out of their cards, and gain an edge over their competitors. This is highly sensitive information and undoubtedly cost them a lot to develop. To open source their drivers would be to give this all away to their competitors, who you can be sure would be the first to look at the code. A highly efficient driver can get the most performance out of the hardware, but it can't make the hardware better than it is. An open-source driver might reveal some coding tricks to make better use of nVidia hardware, but that probably won't help Intel or ATI drivers, and certainly won't help their hardware. Further more, the R&D cost has already been spent, and since they don't sell the closed-source driver, they can't recoup that money except through hardware sales. I can't think of any possible reason why an open-source driver would do anything but increase hardware sales. Intel made this decision, and is now the most widely recommended video hardware I see for Linux. Dell's Ubuntu line uses Intel instead of the nVidia chipsets of their Windows equivalents specifically because they have open-source drivers.

As good as the open source environment is, they do not have the specialists or the resources to develop a cutting-edge product which is, after all, exactly what the video card market is. That point might be relevant if people were trying to create an open-source GPU, because the GPU is complex and cutting-edge. But we aren't asking for that, we're asking for an open-source driver, which only needs to pass instructions back and forth between the Linux kernel and the graphics card. This is a remarkably easier task, and certainly within the ability of open-source development.

Re:A question

[Fross]

I suggest you ask Nouveau whether that is as easy as you make it sound. ;)

Anyone who pays attention to the video market knows that drivers can make or break a card, and often driver updates cause significant performance increases. Look at some of the early Detonator releases, or the alternative drivers released for 3d cards (such as OmegaDrivers). They are not simplistic or simple.

Re:A question

[TheLink]

Yeah.

Maybe it'll eventually be a bit like MP3 audio compression. You try to cut corners for the "resource expensive" stuff in a way that most people are unlikely to notice.

Re:A question

[mhall119]

Nouveau will probably tell you that their biggest hurdle is not writing efficient code, it's reverse engineering the hardware interface. That is why people who want to help them are asked to run the REnouveau utility.

Re:A question

[WilliamSChips]

nVidia really has little reason not to open their source code to the public, unless they are doing something illegal or extremely unethical in their drivers. (Cheating at benchmarks, etc.) Actually, they do have a reason. They would have to remove all the code covered by SGI's patents(which they licence) not to mention all the other companies. If Nouveau gets off the ground and supplants nv they will move their engineers working on nv to Nouveau.

Re:A question

[Anonymous Coward]

Surely we haven't got that many mini-RMSes?

It's only quasi free, semi free, it's the margarine of free, the diet coke of free, just one calorie, not free enough.

Re:A question

[Hatta]

It's the exact same reason that lobbying a congressman (without money) works.

It does? Since when?

Re:A question

[Anonymous Coward]

The code doesn't have to be GPL or any such. They could release it under their own license that specifically states the code can only be used for a driver for nVidia cards. The only thing necessary is the ability to improve the code at will.

On the off chance people don't already know this: that would make it neither Open Source nor Free Software.

Re:A question

[Anonymous Coward]

"The squeaky wheel gets the grease." -> If you whine about something, it will get fixed
"The nail that sticks out gets hammered down." -> If you whine about something, you're made to shut up

Still not seeing how they might be somewhat opposite?

Re:A question

[Anonymous Coward]

That small team at nVidia is 200 people, and their hiring standards are quite tough to meet. They probably won't get beaten by open source devs.

Re:A question

[Dunghopper]

+10 for stating the obvious.

You're right: Linux has a much smaller market share, and we can understand that less resources will be devoted to developing, maintaining, and supporting Linux drivers. We're (reluctantly) o.k. with that.

But, if they're going to bother developing linux drivers at all, I don't think it is too much to ask for the source, so we can do the maintaining and supporting ourselves.

Re:A question

[PitaBred]

NVIDIA can (and does) say no. We're entitled to ask them to support our ideology just as much as they're entitled to say "Ummm, no". Voting with your wallet does absolutely no good unless the company you're voting against knows WHY they aren't getting the money.

Re:A question

[fucksl4shd0t]

Actually, the trend since the late 70s by software development companies has been to remove choice, and if we don't stand up and resist (and even if we do, it might still happen), all choice will be gone. Considering the amount of devices that we depend on every day that are comprised of software + hardware, do you really want to live in a world where some corporate suits get to decide what choices you have and what choices you don't have?

Why is your free market only available to companies? Don't customers have any freedom in your market? We ask for products to be produced under our terms in exchange for our money. Why is that so unreasonable? "It's a free market" has become synonymous with "I have to bend over and let [insert your favorite company here] do whatever they want to me", and that's just plain sick.

Re:A question

[VisezTrance]

The code doesn't have to be GPL or any such. They could release it under their own license that specifically states the code can only be used for a driver for nVidia cards. The only thing necessary is the ability to improve the code at will. (I think they would find it advantageous to go to GPL later, but that's another discussion.) You are right, but I can't help thinking, how many distros could ship those drivers by default ? will I have to download them from a third party repository (deja vu..) will the licence be permisive enough so that the devs consider the code can be merged into the kernel ?

Re:A question

[Aladrin]

The 'only for nVidia' license wouldn't pass any of those things, no... But then, their current drivers don't either. Big corporations rarely make major moves that don't appear to gain them much. That's how nVidia sees this right now.

If they can be convinced to take this little step, and the community shows their support and makes their driver ever better, they'll have reason to seriously consider a more permissive license.

Idealism is great, but you have to be careful how you present things. If you just say 'open source or nothing!' then you'll get nothing. If you say 'see what this step can do for you?' you are much more likely to be persuasive.

Re:A question

[pkphilip]

And why should they listen? Linux users certainly don't form the majority, and so they can't really demand that these manufacturers release the drivers in a specific form?

And by the way, I don't think the fact that Linux users really don't have much of a say is that obvious to the Linux folk. We assume that the manufacturers must respond to our demands because we assume that we matter to them.

Re:A question

[Yfrwlf]

You're right, you always have a choice, and an opinion, and I gave you mine, and you gave me yours, about the things we like and dislike and about the fact that you have a choice. We agree, yay! Now, stick with and help out open source, because there are other benefits like standardization (which is very hard for commercial software to achieve) along with the freedom to control everything. For those companies out there who want a better business model, http://www.opendevelopmentnetwork.org/ is a solution for them that myself and others have been working on. Or at least, thought about working on. It's mostly just a thought at this point due to lack of time to commit to it and the lack of help. ;)

Considering

[Z00L00K]

the fact that the "nv" driver is buggy there is a lot of room left to improve on here.

I experienced a problem with the "nv" driver on my computer with dual 7600GS cards and three displays. It wasn't possible to run all three displays at all with the "nv" driver, but the binary driver from nvidia works. The part that I'm not satisfied with is the need for an alternate driver.

I haven't tried the Nouveau driver, but somebody else may. As I see it, Nvidia should release all information needed to allow others to write suitable drivers. (should apply to all HW manufacturers).

Re:Considering

[Anonymous Coward]

The part that I'm not satisfied with is the need for an alternate driver. Yes and do we really need an alternative operating system. At least my nvidia card works fine with windows...

(awkward silence) ...I'll get me coat.

Re:Considering

[slamb]

[Considering] the fact that the "nv" driver is buggy there is a lot of room left to improve on here.

Definitely. Having a high-quality open source drivers is not just about having 3D support. As is, there's simply a choice between a mostly-good proprietary driver and a horrible, incomplete open source driver.

"nv" driver does not support all cards, has bad artifacts on my GeForce 6150, and fails to use DVI on large video modes with "NV(0): Mode "1920x1200" is larger than BIOS programmed panel size of 1280 x 1024. Removing.". (See bug 3654.) The nvidia people on the bug (look at the email addresses) said that they couldn't fix it with any but the latest GPU. That's obviously false...the proprietary drivers work fine.

nVidia won't fix the open source driver or give anyone else the information to do so, so a reverse-engineering effort is necessary.

Re:Considering

[dpilot]

>nVidia won't fix the open source driver or give anyone else the information to do so, so a reverse-engineering effort is necessary.

Then it's not really open source. Perhaps it meets the letter, but clearly not the spirit.

Re:Considering

[slamb]

Then it's not really open source. Perhaps it meets the letter, but clearly not the spirit.

I agree. You could make a fair argument that they never released the full source code, as the hardware specification should be included.

Re:Considering

[adam.skinner]

The part I'm not satisfied with is hearing people complain that a company offers a linux driver that actually works well! Not only that, the fact that they actively contribute to the open source driver for their hardware simply isn't good enough - it needs to be perfect.

I suggest you learn to be satisfied with less.

someone talk sense into this guy. I'd like ppc

[MrJerryNormandinSir]

I would really like to see a ppc driver! Comeone! And the driver should be written so it will run on the PPC, the new PPC platforms that are emerging, and IBM powerpc boxes like the p4, p5, and p6

COME ON!! PLEASE??????

Re:someone talk sense into this guy. I'd like ppc

[DpakoH]

and PS3 too... This will be best thing in the world probably

Re:someone talk sense into this guy. I'd like ppc

[Anonymous Coward]

And my can opener! Come on! Geesh! _THIS_ would be the best thing in the world. No probably to it!

NVIDIA driver team members use Debian?

[Karellen]

From TFA:

"Across the NVIDIA Linux Graphics Driver team, everyone has their own favorite Linux distribution as their primary desktop: Debian, [...]"

Interesting, given that Debian can't ship their driver.

Oh, I know that none of the driver team will be using a distro-bundled version of the driver anyway, but still...

Re:NVIDIA driver team members use Debian?

[krmt]

None of the major linux distros actually ship their driver. In addition, Debian does have it in non-free, and it's very actively supported there.

Re:NVIDIA driver team members use Debian?

[neersign]

something tells me that the people at Nvidia could care less about the "linux way", and could really care less about the ethics of choosing a distro to match their driver development.

Re:NVIDIA driver team members use Debian?

[Peter La Casse]

I hope they don't develop video drivers on their desktops. It would be horrible to close one's email, web browsers, terminals, IM, IDE, and everything else every time one wanted to test a new driver.

I occasionally compile stuff on my Kubuntu desktop, but not anything that prompts a reboot. That's what test machines are for. Yes it's good to eat your own dogfood, but not when every crash leads to a 10 minute break in concentration while you reboot and start everything up again.

Re:NVIDIA driver team members use Debian?

[Anonymous Coward]

I don't understand why you say that. Just because their driver isn't open source? They have never had anything against linux, they even have a decent linux driver. The windows driver isn't open source, the windows drivers have been known to have bugs. Hell they were late developing a driver for vista for some of their cards.

I think there is a big differance between the linux way and the open source way. If you don't want to show your code to someone it doesn't make them evil, its not like they charge you for their propriatary driver. Not like they go out trying to ruin the open source alternative. I think it's nice to see a large company going out of its way to produce something for linux in the first place.

I do wonder why they don't release the hardwares documentation so that developers could make their own easier, but I could see that being a problem for someone cloning their hardware.

Someone might say that nvidias drivers are a big security hole, and the only way to make it better would be with opensource. Well that might be true, but I guess if your using your computer in a way that it has to be that secure I'm not sure what you need high end 3D graphics for? Computers that are used at an organization that does graphics should be on a tight enough network to begin with.

Re:NVIDIA driver team members use Debian?

[PitaBred]

Ubuntu isn't major? News to me...

What a crazy world we live in

[east coast]

The interview is in English but all the comments are in Czech

Stop the planet. I want to get off!

From now on I vow that when I code the code will be in English but the comments will be in Slovak.

Re:What a crazy world we live in

[Enoxice]

And the readme's are in sanskrit. Expect 6-8 weeks for delivery of the stone tablets.

Re:What a crazy world we live in

[the_arrow]

Well, it's time to switch from hungarian to slovakian notation!

Re:What a crazy world we live in

[Hatta]

Stop the planet. I want to get off!

So the earth moved for you too huh?

Re:What a crazy world we live in

[Tycho]

I have a friend who actually had to try and read and then understand code like that. Only the comments were all in German and he only knows English. The code had something to do with power distribution over power lines, so this was rather important code. After a little over two years he quit rather than move when the company was bought out. Now he works as a manager at a gas station. I think I remember him describing the coding he had to do as being soul crushing.

please help

[Anonymous Coward]

my mom has a myspace page....which is like soooo embarrassing!!!! please troll her into getting rid of it....thanks

                http://www.myspace.com/amandagrashel

alex

Drivers for Syncing

[Doc Ruby]

So far I've never plugged any HW into my Ubuntu PCs without finding at least a proprietary driver that delivers at least the main functions, except an X3d "3DWorld" VGA/glasses system that seems totally orphaned. But I don't use other exotic HW, or even push the limits of the OS with gaming.

What I do lack is sync "drivers" for Evolution. I want to sync across the LAN with apps on Windows. And most important, I want to sync with my SonyEricsson K750 phone. I want to sync my calendars, contacts, and media files, including phone SMS'es as Evolution memos and Evolution email (from a single folder) as some kind of phone text item. I also want to import my old Palm Pilot calendar, contects and memos into Evolution (or the K750, then sync that with Evolution).

Usually OS connections to devices develops slower than app connections, because OS development is slower and has so many other focuses. Evolution seems pretty badly interconnected to other apps, especially on other OS'es, especially on mobile devices.

Or am I just missing a treasure trove? FWIW, the "MultiSync" project is pretty dead, and never worked as easily as a driver (for either the OS or for Evolution).

Re:Drivers for Syncing

[CmdrGravy]

I agree wholeheartedly with you, I haven't found any sensible working method to sync with my phone. If all these people who are just dying to have a crack at the Nvidia open source drivers would work on this in the meantime I'd be very happy.

In the meantime my nvidia graphics card works perfectly on Linux and has done since I bought my first nvidia card 7 years ago so open or closed driver or whatever I think nvidia deserves a lot of credit for that. The last I heard from nvidia they couldn't open source their 3D driver because of licencing issues with some of the technology it uses, I don't think this has changed and if it did I'm sure they'd open source it in a second so all this moaning is pretty pointless in my opinion.

Re:Drivers for Syncing

[Anonymous Coward]

"Or am I just missing a treasure trove"

not sure about that but this may help a little:

I use syncml from evolution to a syncml server (on the evolution machine use "syncevolution")
From the syncml server you can sync to anything that speaks syncml (like a nokia n95, Windows desktop...)
so desktop syncmlserver anything else
The syncml server will usually support other syncing methods as well.

Re:Drivers for Syncing

[Doc Ruby]

That sounds pretty good. Will it sync a phone plugged into a (Linux) PC across the Internet/LAN to the server?

Do you know which Debian/Ubuntu packages contain the syncML plugin for Evolution, and which contain the syncML server? Or where on the Net to find tarfiles/etc?

Re:Drivers for Syncing

[Anonymous Coward]

If all these people who are just dying to have a crack at the Nvidia open source drivers would work on this in the meantime I'd be very happy.

If all of these pansies would just reverse engineer the Windows drivers and build up a state machine to make a complete Linux driver maybe it would actually be done. Laws be damned. Do what it takes. Put up or shut up. This problem has been around for a long time.

nVidia's Buggy Closed-Source Drivers

[BillGatesLoveChild]

One thing that seriously blows about nVidia is their (closed-source) drivers are buggy. e.g. the nv4_disp infinite loop bug. This is year sold, affects many generations of cards, and nVidia *still* haven't fixed it. Fixed it? You can't even talk to them about it. They don't answer any tech support questions: Their forums are user-to-user. If this driver was open source, someone would have fixed it by now. It'd be quicker than the black-box game people have played to fix it.

Read it and weep:
http://forums.nvidia.com/index.php?showtopic=4432& st=80
http://www.christopherjason.com/articles/nvidia-nv 4disp-problem/
http://fileforum.betanews.com/review/950852325/1/v iew
http://forums.2cpu.com/archive/index.php/t-10410.h tml
http://www.tech-forums.net/pc/f77/3d-games-kill-my -pc-need-advice-111501/

I'll share my fix for the nVidia nv4_disp infinite loop bug.

1. Shut down windows.
2. Power off PC.
3. Remove nVidia card.
4. Throw in trash.
5. Install ATI card.

Re:nVidia's Buggy Closed-Source Drivers

[spikeb]

and get even worse drivers! horray!

Re:nVidia's Buggy Closed-Source Drivers

[LubosD]

I have never experienced this on Linux - and this article is about Linux. Moreover, nVidia's Linux team does always respond to received bug reports.

Re:nVidia's Buggy Closed-Source Drivers

[Anonymous Coward]

I find this pretty funny. If you read the stories from yesterday about AMD/ATI Linux driver development, you'd see lots and lots of people complaining about the ATI drivers being buggy, lacking features and being so much slower. At least I can get multiple desktops with Beryl/Compiz working with nVidia. I can't with ATI. I can't even get Xgl working correctly with the binary ATI driver.

Be happy with what nVidia does give us. It is better than the alternatives.

Re:nVidia's Buggy Closed-Source Drivers

[BillGatesLoveChild]

> Moreover, nVidia's Linux team does always respond to received bug reports.

They are accepting bug reports of their Vista drivers too, *but* you too may wake up one day and find support is dropped. Guess they'll do this once the market has critical mass. After that, you're on your own. Well, they actually tell you to talk to your board maker (despite the fact nVidia writes the drivers, not your board maker.)

Re:nVidia's Buggy Closed-Source Drivers

[prelelat]

exactly, I have an ATI card, I promissed myself that I would never buy one because of the linux support. We'll I was doing some work for a couple years that required me to have windows installed. A buddy of mine convinced me to use ATI, I figured they have better drivers now for linux if I ever switch back it won't be a problem. We'll I switched back to linux a year after I was done the work and I regret ever spending the money on this ATI card. It worked nice in windows but even the propriatary drivers have a hard time using two monitors. It just doesn't work right. The NVidia drivers at least worked properly. I don't need a new video card I don't do that much with high end graphics, but I almost want to get an NVidia card because it sucks so freaking bad. The open source drivers are almost better, but they still don't work as good as nvidias.

Frankly throwing out your NVidia card what are you going to get? An intel card, I'm not sure how good the drivers are for that, but really NVidia seems to be the way to go, and just hope that the opensource project advances quickly(either for nvidia or ATI)

Re:nVidia's Buggy Closed-Source Drivers

[Anonymous Coward]

A) He's using Windows. ATI's drivers are fine on Windows.
B) Even under linux, where ATI's drivers suck ass, I've NEVER encountered a bug of that severity.

  I have an old Radeon 9800, bought when it was the new hotness, and I've encountered the following bugs since that time on linux:

  * Drivers are tricky to install (this was fixed long ago)
  * Drivers are slow -- feels like about half the speed of the windows drivers
  * Switching back and forth from X to console sometimes garbles console until reboot (haven't seen it recently)
  * Dual monitor doesn't work very well
  * Penumbra would hang at a specific point until I removed some of the static libraries that shipped with it
        (May be driver related, may not -- I think it was related to a graphical effect, but could be audio, too)
  * Won't do AIGLX

  That's it. None of those bugs are anywhere near the severity of the bug he's described above. The whole system dies immediately on boot? WTF?
  I wasn't even aware of a bug like that until I read his post -- all I've ever heard was that nVidia's drivers were made of sweetness and light and happy dreams about puppies. OTOH, I've had driver-related crashes with an MX440 on a windows box, (later revisions fixed it) so I never really bought into that line anyway.
  Right now, I'll hold off on purchasing any new card, while I wait to see the results from AMD's upcoming unified OpenGL driver. They're rewriting it from the ground up, and it will run the same driver on Vista, XP, and Linux. A good idea, IMO, since all indications are that the linux driver was hastily cobbled together and has severe architectural issues preventing major bugfixes and feature additions.

Re:nVidia's Buggy Closed-Source Drivers

[AaronW]

I second this. I curse the drivers for my ATI card daily at work. Xemacs leaves crap all over the place with ATI, Google Earth hangs, and the 3-D on the X1300 is slower than molasses, much slower than a G-force2! Even the mouse cursor frequently gets corrupted with Xinerama.

I can watch it repainting the screen with the screensaver!

Give me nVidia any day, even an old card over the proprietary crappy fglrx drivers.

The other thing is, when I have experienced problems with the nVidia drives and reported them, I have always been contacted back by nVidia in an attempt to resolve the problem. I have yet to even get a single response back from ATI/AMD even when they update the status of the bugs I've reported.

WRT Linux, ATI cards are only good for the dumpster until they get their act together on drivers. And no, I only need working 2-D support, and the open-source drivers don't work with this card at all.

-Aaron

Eating your own dog food

[should_be_linear]

I even know Windows driver engineers at NVIDIA that primarily use Linux on their personal machines at home.

Well, that explains lot of issues with my DirectX setup.

Noticeable absence

[redelm]

I read TFA fairly closely, as usual looking as much for what could have been said but wasn't. iReading tea leaves. First, NVidia did not complain they had licenced technology from others which they couldn't "open source". It would have been a great excuse, but since they didn't use it, it doesn't apply.

Second, while they certainly want to be seen as supporting Linux, they really believe their closed-source drivers give them some source of competitive advantage. That's either in clever code or what the coding reveals about the internal organization of their GPU hardware. It would have been relatively easy and palatable to say: "We'd like to release full GPU asm specs and code, but believe this will help our competitors design better hardware. So we can only provide APIs." They didn't say this, so I think they consider their actual driver code to be very clever (main competitive advantage). No such secret will last.

Yes, I know there are many other explanations for "negative knowledge" -- things that didn't happen. But when they could have and would have been easy, perhaps we need to wonder why they didn't.

Re:Noticeable absence

[Anonymous Coward]

My guess is the lawyers said "we may be potentially violating N patents" (N being large) and they have to keep the drivers closed to avoid handing out ammunition for lawsuits.

Re:Noticeable absence

[Anonymous Coward]

No such secret will last.

It might only need to last until after the successor products
have been released and marketed for a few months, although that
would depend on what critical(ly proprietary) architectural
features the successor design inherits from its predecessor.

However, there's always the risk that the customer base would have,
by then, grown disenchanted with the vendor's attitude towards
open source and migrated to their competitors' products.

Re:Noticeable absence

[Anonymous Coward]

We don't want their code. Let me say it again. We don't want their code. We want specifications on how to interface to the device. We want to be able to talk to the GPU like we are able to talk to the CPU (or any other sane piece of hardware).

Why do GPU manufacturers think they have more IP stake then CPU manufacturers? Imagine a CPU manufacturer saying that the CPUs opcodes are proprietary that they don't want to release for competitive advantage reasons. Just use our compiler they say. That would be absurd.

Re:Noticeable absence

[Apreche]

In my estimation, I do think that the coding of the NVidia drivers must be very clever. Even if you compare the Windows Nvidia drivers to the ATi drivers just from a user's perspective, the NVidia driver blows ATi out of the water. It is more stable, more usable, more reliable, and more compatible in every possible way. In fact, I bet what they are doing is so clever, that even if they released the specs for all their hardware, the open source driver would remain inferior for a very long time.

Re:Noticeable absence

[darkwhite]

No such secret will last. Not to support nVidia's closed source policy, but modern technology is founded upon secrets that don't last. Companies develop industrial procedures, capitalize on them for a decade or less (much less in the case of the computer industry), then move on, while the secrets that generated millions of $ in revenue either disappear or become public knowledge mulled over and taught by university professors.

nVidia's stance may or may not be reasonable, but consider the fact that their main competitor - ATI - has been lagging significantly behind in all areas of R&D for the past two years. I don't know much about video driver and GPU architecture, but it's possible that with open drivers ATI would gain significant insight and wouldn't be so badly behind (which would be good for consumers, bad for nVidia).

Re:Noticeable absence

[Anonymous Coward]

Yes but, the absence of evidence is not evidence of absence!

Re:Noticeable absence

[Anonymous Coward]

Let's assume (for the sake of argument) that you're right: the nVidia proprietary drivers are wonderful and these open source losers will never be able to duplicate their perfection. How does that change the cost/benefit analysis? Let's examine it from their perspective:

Benefit: nVidia gains good will for holding up their end of the bargain and respect for having obviously superior drivers.

Cost: ...?

I'll still switch..

[pjr.cc]

ATI said they'll go OSS with their drivers and if they do i'll switch away from nvidia and be happy to do so given that article.

Re:I'll still switch..

[Dan Ost]

When did they say that they were going to opensource their drivers?

Re:I'll still switch..

[stecoop]

Go ahead and make the Switch to Intel video cards, the support for Linux is currently the best avalible.

Re:I'll still switch..

[eht]

Even if they did say that, it is so cute he believes it.

Re:I'll still switch..

[pjr.cc]

Even if they did say that, it is so cute he believes it.

I never said i believed it, just that i would switch if i did. I really like nvidia cards to be totally honest, but here was the article i was refering to. Unless I've miss-read that, then it says AMD will OSS the ATI drivers.

When it comes to hardware, i'll happily support any vendor that delivers something usable in the linux community.

Re:I'll still switch..

[Anonymous Coward]

Which is exactly why I support Intel and intend to use all of their hardware on machines where I don't need the extra performance of a discreet GPU. If ATI/AMD truly follows through I will most certainly switch to that platform.

This seems like a no brainer for AMD as it will give them a competitive advantage. It isn't all about absolute performance, at least for many of us, but a package deal that has to be attractive. Considering my limited use of 3D under Linux it is not like I'm real demanding performance wise, but even those that demand performance ought to have some concept of the goodness of accessibility.

Dave

Re:I'll still switch..

[Dan Ost]

I'd love to see it happen.

Here's hoping.

Re:I'll still switch..

[Anonymous Coward]

Err, the ATI drivers for linux are WORSE than the nvidia ones...

Re:I'll still switch..

[Anonymous Coward]

Their VP of Sales said they were going to have an "open driver" by the end of the year. He may have been letting the cat out of the bag about AMD's near-future plans, or he may have meant "OpenGL driver" instead.

  The latter's a strong possibility, AMD has already announced a ground-up rewrite to fix the longstanding issues with ATI and OpenGL -- both their lousy linux driver and poor OpenGL performance under Windows, as compared with DirectX performance. This will also entail a single, unified driver between all supported operating systems.
  Even so, I wouldn't count out the possibility that he really did mean "open driver" -- AMD considers itself a hardware company, not a software company, and with Intel already out there, they may want to compete in that arena.

  I know that nobody from ATI has contacted the Xorg driver teams yet, but that wouldn't surprise me. It took ATI/AMD about five months just to approve and release a set of slides describing their new driver development schedule, so I imagine any "let's open-source the driver" plan will spend at least that long in lawyer-land getting reviewed and revised and re-reviewed before any real action gets underway. Publicly traded companies have all kinds of rules a regulations that they have to comply with to verify that they're not trying to manipulate the stock market, and that stuff takes time.
  (Not that I'm ranting against regulations here. In this case, the alternative is "Why don't we simply trust them not to cheat people?" and the answer is "Because then they'll cheat people.")

There's another interview with Nvidia's...

[Saint Aardvark]

...Director of Unix Development, Andy Ritger; the BSDTalk podcast interviewed him and Christian Zander last year about NVidia's support for the BSDs.

Isn't that obvious?

[Anonymous Coward]

There are two reasons:

First, drivers are the half of the product. They are as important as the actual hardware. Open sourcing drivers would be equivalent to releasing VHDL code for their chips. It would give a huge advantage to their competitors who could utilize the same tricks to improve performance of their products.

Second, not all code in NVidia's drivers is owned by NVidia. They have licensed some stuff from other companies and they can't open source that code even if they wanted. Keep in mind that NVidia isn't developing Linux drivers separately, most (>90%) of NVidia's driver code is common across all OSes.

Re:Isn't that obvious?

[QuantumG]

How the hell do you know? Is this anything other than more speculation?

Until NVIDIA tells us why they cant or wont open source their drivers, we simply don't know.

Stop speculating, it's not productive.

Just a thought...

[Ekhymosis]

Perhaps I'm echoing something that has been said before, but please bear with me. If the companies (Nvidia and ATI) are worried that releasing specs to the open source community will violate patents, why not make the driver more modularized so that they can release everything that is not patented and have the patented parts be libraries/plugins/whatever? This would allow the community to have better information leading to relatively more stable drivers (not 100% due to the binary 'plugins') instead of reverse engineering which is not always a delightful thing to do. I hope I'm making sense here...=p

Re:Just a thought...

[Anonymous Coward]

Honestly? Because it requires additional work.

I know, I know, rules of proper design, etc, but open-source is still last on the plate to getting something out there that will sell to the customers and bring in money. I mean, seriously, what is management going to insist engineers apply more effort towards:

(A) Creating a well-designed, cleanly modular system driver, one that helps maintain a well-maintained codebase and would also allow for easy open-sourcing?
(B) Tweaking for months to squeak out a few more points in 3DMark/a few more fps in Doom 3?

A different type of driver

[simong]

There is a difference between drivers for graphics cards and 'drivers' for synchronisation software. In the synchronisation world there is a standard, SyncML/OMA Data Synchronisation, which has been adopted by M$ and most of the major phone/PDA companies, but hasn't found its way into Evolution or other OS desktop systems yet to my knowledge, even though there is a server for it. This topic is talking about closed source hardware drivers though, and cable linked PDAs usually use some kind of serial port, so there's plenty of support for that.

nVidia PS3/Linux Driver?

[Doc Ruby]

How about a driver for Linux on the PlayStation 3, which would let the PS3 RSX chip actually work for Linux apps?

Right now PS3 Linux runs all display processing on the PPC core on the Cell, which needs to do a lot of other processing to keep the complex Cell going. Meanwhile there's an RSX chip that runs at 1.8TeraFLOPS, dwarfing even the Cell's 0.2TFLOPS. But Sony's Hypervisor virtualization layer that runs Linux hides the RSX from Linux. However, the RSX is exposed in some API, otherwise PS3 Linux wouldn't display on the HDMI port out of the PS3, and sound probably wouldn't work (probably also running on the RSX somehow).

Sony doesn't want the RSX exposed to Linux apps, because then Linux apps could compete with Sony-licensed games (without paying Sony the royalty that even subsidizes over 25% of the PS3 purchase price). But can't nVidia release a driver, or some kind of specs, that expose a 2D API for running X desktops? Sony's money all comes from 3D games.

Or maybe someone else has a way.

Re:nVidia PS3/Linux Driver?

[Anonymous Coward]

Sony also don't want the RSX exposed because it would probably be possible to circumvent the hypervisor that way. Because the RSX is a last-minute off-the-shelf drop-in component it doesn't have security built in like the rest of the system.

Re:nVidia PS3/Linux Driver?

[Doc Ruby]

The RSX is a part unique to the PS3 designed jointly by Sony and nVidia.

What makes you think it lacks the security the rest of the PS3 has? What makes you think exposing RSX functions in a driver would expose holes in the Hypervisor, other than the access to the RSX that the driver exposes by design?

Re:nVidia PS3/Linux Driver?

[Anonymous Coward]

Sony are not concerned about competing with Linux games on the PS3. The Linux gaming market is near nil, and will probably remain like that. The number of real Linux users on the PS3 that aren't playing with cell development is significantly less.

There is another reason Linux is crippled on the PS3. What that is, I have no idea. Sony are extremely quiet about it. I suspect they knew is was pretty easy to get Linux installed due to the existing PPC distros, and fed us a BS line about the PS3 running Linux to gain some small tech cred.

Anyone who has actually used Linux on the PS3 will know it's pretty damn useless as it stands. RSX lock-out and fsck all memory.

Re:nVidia PS3/Linux Driver?

[Anonymous Coward]

Sony originally planned to use a second Cell CPU for graphics, but at some point pretty late in the design cycle realized that it wouldn't work (too slow, developers wouldn't want to write the rendering code) so they went to nVidia instead. Due to the time constraint there was no way for nVidia to design a proper graphics system from scratch, so instead they modified one of their existing cores.

The security aspect comes from the fact that the RSX can DMA to and from anywhere in host memory, meaning the hypervisor can trivially be broken. AFAIK there is no IOMMU or similar prevention mechanism in place which most likely would have been the case if the RSX would have been in the design from the start. If you think that something like a binary driver would help stop this you have no idea to what lengths console hackers and especially pirates will go.

Re:nVidia PS3/Linux Driver?

[Doc Ruby]

I don't think you're reading my post properly.

Sony wouldn't be competing with "Linux games" in a "Linux gaming market". Sony would be competing with bootable CD/DVD/BD games that run on PS3, just like Sony games. The gamer wouldn't know it's a "Linux game", it would just be a game. That would be cheaper than the Sony games, because it doesn't include the Sony license (among other efficiencies).

Of course that scenario depends on the RSX working. So I don't know why you're talking about "RSX lock-out", which my post to which you replied postulated as working, the basis for competing with Sony-licensed games. Perfect circle, useless argument.

BTW, Sony has sponsored Linux on PS3 itself, not just feeding us a line about it working.

Re:nVidia PS3/Linux Driver?

[Anonymous Coward]

Can you make 2D games for PS3/Linux? Or are Linux programs locked out of graphics completely? If a game was sufficiently good, it might not matter if it wasn't 3D...

(I'm not saying I have such a game... just curious)

Re:nVidia PS3/Linux Driver?

[jonwil]

linux apps can't compete with PS3 licensed games. In order to run any linux software you need to install the PS3 otheros loader first. No games company is going to want to cut down the market by forcing people to install that loader.

I suspect the real desire is to stop things like Second Life (which competes directly with Sony's new online virtual world offering) being ported to PS3 linux. The people who are likely to use things like Second Life are more likely to have the know-how to install the otheros loader and such.

Re:nVidia PS3/Linux Driver?

[Anonymous Coward]

You do have a 2D framebuffer, so games are definitely possible. I bet you could even make somewhat decent 3D graphics by dedicating a couple of SPUs to rendering.

Re:nVidia PS3/Linux Driver?

[NereusRen]

How about a driver for Linux on the PlayStation 3, which would let the PS3 RSX chip actually work for Linux apps?

I know reading the article is considered some sort of terrible burden around here, but if you're going to ask a specific question, you might as well see if the interviewer already asked it for you:

4) Do you expect this situation [x86-only support] to change in case that Sony allows Linux to access the RSX chip on its PowerPC-based PlayStation 3?
We have no current plans to port our Linux driver to PS3.

Re:nVidia PS3/Linux Driver?

[Doc Ruby]

Yes, except that I'm specifically asking for a new driver for the RSX, ported from XMB to Linux. Ritger said they wouldn't port their Linux driver, which wasn't even a subject because no one ever said such a thing existed already.

More to the point, continuously asking nVidia and Sony for a Linux RSX driver, by a whole community, can pressure "no current plans" into "we've now made a plan". Which could accommodate the specific compromise I mentioned, without threatening Sony's essential interests.

I know understanding the issue and the process that published interviews and feedback play in affecting the issue is too sophisticated for Slashdot readers, but here's some help. We want better platforms, not just official pablum from vendors.

Re:nVidia PS3/Linux Driver?

[Doc Ruby]

If games cost $20, or $5-10 or even $40 instead of Sony's $50-80, then you'd see a lot of people installing the loader. Especially if it were packaged as a single/few clicks, dead simple, even if that were on a separate medium. If the process is improved so the same CD/DVD/BD/memstick could install the loader, then install the game (or game loader, etc) to the HD, so it's really simple, then the market would be substantial. The larger market for cheaper games, especially the more diverse games, perhaps enhanced by more easy copying, could be bigger than the market for expensive Sony games (especially once the cheap competition reduces the market for expensive ones), even if there's a barrier to entry in installing the loader.

The substantial market for Xbox and other console hacks/mods, that the console vendors spend lots of time, money and bad PR aggressively hunting to extinction, demonstrates that Linux games, even requiring a (one time) upgrade with some complexity, would be legitimate competition to XMB games - if the price is right.

Remember that competition doesn't have to threaten a larger market share to be worth destroying. Just threaten to take a larger market share than the cost of destroying it consumes.

Re:nVidia PS3/Linux Driver?

[Doc Ruby]

You can make 2D, 3D, whatever graphics you want: as long as it runs on the existing OS support for the PPC, which just writes to a framebuffer. That PPC is a little faster than most x86 chips, but the x86 chips all offload their graphics to accelerated "VGA" (and OpenGL etc) chips. While the PPC is also busy doing a lot of other stuff to support the Cell's complex functions.

When someone ports OpenGL to the Cell's SPU(s), that will make it faster than any x86/VGA/OpenGL machine today. But that's time consuming, complex, and the RSX already does all that, if the Hypervisor is just opened.

Re:nVidia PS3/Linux Driver?

[MikeBabcock]

I'm not holding my breath for Mesa to be ported to an SPU but it would be interesting I must say.

The line is blurred

[Anonymous Coward]

At one time, there was a very clearly defined line between the graphics hardware and the driver software. The hardware was sold, the drivers were given away. The trade secrets were in the hardware. The driver just programmed registers and set up data transfers. In this environment, it is easy for someone like Stallman to draw a distinction between hardware and software.

However, GPUs are reconfigurable computers. Some of the rendering hardware is actually being built dynamically out of reconfigurable components to match the current tasks. This is very powerful, but it also means that the drivers contain partial descriptions of the hardware. The line between the two components is indistinct - where are the trade secrets of the hardware construction? Answer - they are spread between the driver programs and the GPU ASIC design.

It's exactly the same situation that we hit with the Winmodems, before DSL killed them off. Some wireless chipsets have this problem too, and at one point there were "winprinters" too. We're going to hit this again in the future, because CPUs will become more like reconfigurable computers, and there will be binary blobs to gain access to the high speed features. It's not the end of free software, but it is a major challenge for us all.

cool!

[DavoMan]

Cooool. Good to see that Direct3D isn't on -everything-.

Re:cool!

[goaty_the_flying_sho]

It's not so bad, my wine direct3d layer is faster than Windows' implementation now.

If it's their patent

[Anonymous Coward]

then GPLing it means ATI can't use the patents unless they *GPL* their driver. And if *their* driver also includes patents, then these patents are available to NVidia for their GPL driver.

If it's someone else's patent, then let us know whose and we can tell them the same thing.

GPLing code implementing a patent doesn't help a competitor get a leg-up.

BSDing it would.

Video Acceleration Suppport

[tji]

As a Myth user, trying to improve my HDTV output, I'm interested in the video acceleration capabilties of their components. NVidia provides some basic MPEG2 acceleration support (XvMC) but it doesn't seem to be too robust, there are tons of problem reports on the Myth mailing list. This is one area that could _really_ benefit from open source access to the drivers. But, we know how likely that is.

From TFA, regarding video accell:

Q: ... People seem to be especially interested in better 2D acceleration, hardware MPEG-4/H264 acceleration ...

A: (Nothing Related to video acceleration)

So, continued lack of suppport for all their hardware Video Acceleration capabilities in Linux. The only possibility it seems was touched on in the next question:

Q2) People would like to use their powerful GPUs for their own processing intensive applications (eg. video compression or math formulas computation). Can we expect some SDK to make this happen?

A: For NVIDIA GPU programmability, there are several options: for graphics-oriented applications within the context of OpenGL, NVIDIA supports Cg and GLSL for higher-level programming

GLSL could theoretically be used to accelerate video decoding. But, that's quite an undertaking, and is only supported on the newest chips.

Instead, I think the way to go is with Intel integrated GPUs. They don't have MPEG acceleration yet, but they are working on APIs and drivers to greatly improve this (the hardware has very good video support). I'm hoping Intel GPUs quickly become the far and away best option for Myth users.. let open source prevail.

Re:Video Acceleration Suppport

[kimvette]

Instead, I think the way to go is with Intel integrated GPUs. They don't have MPEG acceleration yet, but they are working on APIs and drivers to greatly improve this (the hardware has very good video support). I'm hoping Intel GPUs quickly become the far and away best option for Myth users.. let open source prevail.

Intel open source drivers are fine and all but before I abandon Nvidia for Intel I need:

  - Dual dual-link DVI ports (the nvidia cards we buy have these)
  - HDTV & S-video ports
  - dedicated video ram
  - in a discrete card (are there add-in video cards based on Intel GPUs?) - if I need to upgrade the video, or replace a bad video card, why the hell should I have to upgrade the board (and processor, and CPU) all at the same time?

I, for one, am hoping that AMD forces ATI to open up their drivers. THAT will create a ripple effect through the entire industry. Nvidia will have no choice but to follow suit.

Re:Video Acceleration Suppport

[raw-sewage]

Q: ... People seem to be especially interested in better 2D acceleration, hardware MPEG-4/H264 acceleration ...

A: (Nothing Related to video acceleration)

I'm one of those people who are interested in better 2D acceleration and hardware MPEG-4/H264 acceleration. I'm a MythTV user. Nvidia's hardware does have support for such acceleration; they call it Purevideo. Unfortunately, the Linux drivers don't support it. Supposedly, it will be supported "in the future". But it's really frustrating. The new 7050PV chipset would be perfect for a high-definition, relatively low-powered (meaning quiet) MythTV system. But alas, the HD decoding features of that chip aren't (yet) supported in Linux, and therefore you still need a beefy processor to do HD in MythTV.

Instead, I think the way to go is with Intel integrated GPUs. They don't have MPEG acceleration yet, but they are working on APIs and drivers to greatly improve this (the hardware has very good video support). I'm hoping Intel GPUs quickly become the far and away best option for Myth users.. let open source prevail.

I wish Intel would release a standalone video card. This would be perfect for my workstation: an enthusiast motherboard for overclocking, but an ordinary video card (since I don't game on this machine). The problem is, the best overclocking motherboards rarely have onboard video.

Re:Video Acceleration Suppport

[Anonymous Coward]

- Dual dual-link DVI ports (the nvidia cards we buy have these)

Okay, if you're driving an extremely hi-res display. For video use, my max is 1080p, which can be done via standard VGA or DVI.

    - HDTV & S-video ports

The X3000 supports VGA, DVI, and HDMI. I'm not sure about the opposite end of the scale, S-Video.. If it's not supported on the integrated video, a cheap PCI/PCIe card could be added. But, who cares about video acceleration when doing lo-res S-Video?

    - dedicated video ram

Why? If the GPU does the job with system RAM, why not use it, especially as cheap as ram is now. I just put 4GB of RAM in my linux box for ~ $130. Video RAM is an issue for 3D gaming performance. If that is what you want, DO NOT go with an integrated GPU. If you want good video performance (e.g. a MythTV box) integrated GPUs are the best solution.

    - in a discrete card (are there add-in video cards based on Intel GPUs?) - if I need to upgrade the video, or replace a bad video card, why the hell should I have to upgrade the board (and processor, and CPU) all at the same time?

Replace a bad video card (chip)? With an integrated GPU, that's basically saying the south-bridge fails.. I've never seen that happen, and I would guess the odds of that are pretty low, if it did happen your whole board is unusable. The boards all have PCIe x16 slots too, so you can always upgrade to an add-in card if desired. But, again, the upgrade cycle is typically for gamers wanting 3D frame rate boosts, not video playback, which was the original topic.

Re:Video Acceleration Suppport

[sbryant]

I'm one of those people who are interested in better 2D acceleration and hardware MPEG-4/H264 acceleration. I'm a MythTV user. Nvidia's hardware does have support for such acceleration; they call it Purevideo. Unfortunately, the Linux drivers don't support it.

Well, the "nv" driver doesn't support it, but Nvidia's own driver does ("it" being the XvMC extension in Xorg).

Given that you're a MythTV user, here's a link to their Wiki, which contains a page about XvMC support.

The new 7050PV [nvidia.com] chipset would be perfect for a high-definition ...

I've not tried it, but the 7300GS just came out a month ago, and I happened to be upgrading so I got one - for 40 Euros; it's passively cooled, and so far has been working really well in TwinView mode on my Linux box. I use Kaffeine (Xine-based) rather than MythTV, but it's very smooth. I'm sure MythTV would be the same.

I wish Intel would release a standalone video card.

BTW, I'm using a machine with an integrated Intel VGA at work. It's OK - has open source drivers, and 3D acceleration, but it uses system memory, only has one output (not DVI either), and doesn't always behave itself. Personally, I much prefer the 7300. Oh yeah - the Intel driver supports XvMC too.

-- Steve

Re:Video Acceleration Suppport

[tji]

> ... The new 7050PV chipset would be perfect for a high-definition ...

Yeah, they have all kinds of great hardware features, none of which are available in Linux. As a long-time HDTV/HTPC user, I've bought my last piece of hardware based on advertised capabilities not yet available in Linux. As soon as there are Linux drivers to use it, I'll buy it, short of that I'm not interested. Even when they add new features to their drivers, I'll still be hesitant. NVidia's XvMC on their X86-64 drivers had a known initialization bug for over two years that caused me some headaches. Closed source drivers suck.

> I wish Intel would release a standalone video card. This would be perfect for my workstation: an enthusiast motherboard for overclocking, but an ordinary video card (since I don't game on this machine). The problem is, the best overclocking motherboards rarely have onboard video.

Yup, Intel's never gonna do an overclocker board. But, I value stability a lot more than a couple percentage performance gain (or $20 saved on buying the cheaper CPU model). I can see overclocking for gamers, but video processing works at a fixed rate, e.g. 59.94 frames per second, cranking up the CPU speed isn't going to matter unless you can't keep up with the frame rate at the normal clock speed.

There have been some rumblings about intel doing an add-in card. But, I kinda doubt it.. that's a crowded market, where they don't want to piss off their good partners like NVidia. But, as a completely non-gamer, I'm happy with the integrated GPU.. low power, low heat, low noise(no fans), no additional card in the PC case. If/when Intel enables the video acceleration hardware in Linux, I'll be first in line to get a board & build my new MythTV frontend. Their open source drivers, and video acceleration API effort are definitely on the right track.

Re:Video Acceleration Suppport

[raw-sewage]

I'm one of those people who are interested in better 2D acceleration and hardware MPEG-4/H264 acceleration. I'm a MythTV user. Nvidia's hardware does have support for such acceleration; they call it Purevideo. Unfortunately, the Linux drivers don't support it.

Well, the "nv" driver doesn't support it, but Nvidia's own driver does ("it" being the XvMC extension in Xorg). Given that you're a MythTV user, here's a link to their Wiki, which contains a page about XvMC support.

XvMC only supports MPEG-2 (e.g. what standard-def DVD movies use and over-the-air HD (in the USA anyway). After bending over backwards and other headaches, I finally got XvMC working on my MythTV, only to find that the video quality was still pretty lousy. Maybe there was still some tweaking left to be done, but search for "xvmc" on the mythtv user's mailing list archive, and you'll find an overwhelming amount of people who are also struggling with it. I think a lot of people do like me, and just concede to getting a fast dual-core processor.

Re:Video Acceleration Suppport

[the_greywolf]

- in a discrete card (are there add-in video cards based on Intel GPUs?)

Believe me, wait until they get their X3000 series right. The last Intel card in my memory was the i740 - and it made sucking donkey balls look like a stately profession. It claimed 3D support, but was barely capable as a 2D rasterizer. I still have mine somewhere.

Re:Video Acceleration Suppport

[PenGun]

-vo xvmc:bobdeint:queue -vc ffmpeg12mc

How about a compromise?

[TheVelvetFlamebait]

I'm thinking releasing the source in seriously obfuscated c with no descriptive identifiers. Or, there's always the old uncommented assembly.

Hey now that's funny

[BillGatesLoveChild]

After I hit the [send] key my trusty nVidia just did its BSOD infinite loop bug. Fourth time today. Getting a new ATI card tomorrow. Stay away from nVidia and don't buy their dulcet cooing: They're bad news.

Re:Hey now that's funny

[idonthack]

If people everywhere were having full-blown BSODs as fast as you are, (fourth time today, it's 10 AM) I'm sure they would fix it. Of course I'm betting you're in a similar time zone.
 
Anyways, are you entirely sure it's not defective hardware (not necessarily your video card) or a corrupted driver? Those seem more likely candidates than a consistant driver bug. Especially when the first post you linked to says

so i got the 91.28 drivers and they finally worked, now i can play games like i did before

With a bit more looking, it turns out at least half the pages you linked to have solutions. I stopped reading after that. If you're claiming it's never been fixed yet there's step-by-step solutions in all of the "evidence" you linked to, you either cannot read (unlikely, you can write) or you are trolling.

Re:Hey now that's funny

[BillGatesLoveChild]

> With a bit more looking, it turns out at least half the pages you linked to have solutions. I stopped reading after that.
> If you're claiming it's never been fixed yet there's step-by-step solutions in all of the "evidence" you linked to,
> you either cannot read (unlikely, you can write) or you are trolling.

And you're astroturfing. Here are some comments from your "satisfied customers" on those very same pages:

1. "oh BOY thats a 115,666 VIEWS!, seems to me nvdia "Fixed" thumbsdown.gif the problem! no.gif
2. "If I were you, I'd go with an ATI, Nvidia is full of crap.
3. "Hello infinite loopers! This weekend I was also afflicted by the curse of the nv4_disp infinite loop, and it was initially reassuring to find so many tips on potential workarounds on this board. But, after trying various recommendations e.g. the nv4loopfix.zip patch, upgrading drivers, and playing with BIOS settings there was no improvement - the PC kept blue screening after only 1 minute uptime.
4. "Got the problem with 6600gt(pcie), (Flickker, crash and so on) Ps. Return the card if ya can. Pps. Freaking nvidia, make better cards/drivers!
5." Hey guys you seem to have the same BSoD as me. But the thing is I have a GeForce 8800GTS - and I installed the latest Nvidia Drivers (97.92_Forceware Winxp_eng) but the probleem keeps persisting.
6. "Hey Ive been having this problem for quite a while,, Im 100% stable in games I can play Oblivion for 10 hours straight withought a problem,, but as soon as I start surfing the web; usually when im using the scrool wheel I get an infinite Loop BSOD, every time? I recently installed the new 93.71 drivers and they worked great for like 2 weeks no more blue screens, but tonight they started up again .. ?????
7. "My problem is with the GE Force FX 5200 too. I had the same problem...Blue Screen Of Death!
8. "just had the problem again this a.m. while using google earth. machine froze, gave me the bad nv_4 message.
9. "I've got two issues. Dell Dim 8300 is about 2 years old but the symptoms are less than 5 days old. 1 - Video screen begins to freeze when my kid is playing Madden 07 2 - nv4_disp error appears before XP finished booting.
10. "Danza, welcome to the club of WinXP + Det 4 = nv4_disp loopout.
11. "My problem, is that the driver seems to go into some sort of loop when initializing either an OpenGL or Direct3D game. I get the BSOD with nv4_disp.dll as the named culprit.
12. "Well I have sent off emails to VIA, Microsoft, and Nvidia, but of course no comment from anybody. I wish some big website would grab this thread and run with it.....force the manufacturers to take a look at the problem and at the very least aknowledge a problem does indeed exist."

Note 12: I've tried contacting them too. nVidia don't want to know about this bug.

Linux-certified hardware

[athloi]

Windows certifies hardware, and Apple makes it clear what they support. Could it be useful for an agency of Linux developers to certify hardware that is open (standards released so drivers can be written) and well-designed enough to support the rigors of a "UNIX-like" OS?

I do not know the answer to this one. My inner four-year-old anarchist is leery of certification in anything, but even something as simple as a list of supported hardware like BSD does, with the requirement that its standards be open so drivers can be developed, might help companies market to Linux users (1 in 10 users, by my estimate) and help Linux users get their market share behind a few quality products so they can stand up and be counted.

Just an idea. Feel free to mod -1, this guy's an idealistic moron.

Re:Linux-certified hardware

[Anonymous Coward]

I'm not sure it's even logistically possible to have a "Linux Certified" program. However, I'm really pleased to see major companies like Dell offer Linux out of the box. Although it's not the same as "certified," at least you know they've put some effort into making sure the hardware and drivers all work. Anybody who's spent N number of hours trying to get a piece of hardware working in Linux knows that I'm talking about.

Blah

[nrgy]

I hate seeing these kind of topics because they always turn into a flame war. I've been using linux as my only operating system for four years straight now and I could realy give a rats ass about whether a driver or program is open source or not. Its pretty sad that you have this vocal hippy crowd that are like children and if something good doesn't have it's source code available then its the work of Satan. Now I've never had a single problem with Nvidias linux drivers but that being said I'm sure their are some. Please stfu with the "See if it were open source the bugs would of already been fixed" shit, if a bug were as simple to fix as you try and act like it is then don't you think Nvidia would already have done it? Like I said really don't give a damn about open or closed because regardless of which of the two it is you WILL have bugs. As much as you would like to think open source is a mega god and suddenly turns a program or driver into a piece of bug free code because so many people will be able to see it, well I'm sorry to say your wrong. It's pretty sad now a days that just supporting linux with a driver or port of an application isn't enough, no the hippies also want access to your bread and butter to. Every time linux gains an inch these zealots run out asking for 3 more, well you can't have it all so fuckn deal with it. I've come to realize this vocal "ZOMGBBSAUCE EVERYTHING HAS TO BE OPEN SOURCE" crowd is nothing more then people that can't fully do something themselves so they want to milk off of someone else's work and have them do it for them. Mod me flaimbait all you want I could care less. I'm a linux user not because I follow some open source screwed up religion, I like the idea of open source but at the same time I'm sane enough to realize its not the only option. I don't play in the middle ages like some of you do, forcing people to convert to your way just because they play in the same sandbox.

Re:Blah

[Anonymous Coward]

How long have you been off your medication?

Re:Blah

[michael83r]

hahhaha... I too have never had a problem with nvida drivers on linux. Although on the other hand i can never get the ati proprietory drivers to work..

Please

[Anonymous Coward]

Oh please, it's a slightly modified GeForce 7800 chip.

Re:Please

[Doc Ruby]

OK, smart AC guy, tell me how it's modified, in terms of specs for modifying a 7800 driver to make it work on PS3 Linux.

Re:Please

[Anonymous Coward]

Well, you can start by practicing your binary translation skills..

[ot] What a crazy world we live in

[wild_berry]

That's fast. The last chap was up the mountain getting them for 40 days but then spent 40 years wandering the desert...

Andy Ritger

[Anonymous Coward]

Does anyone else remember when Andy used to be Conan O'brien's sidekick? I always said to myself, "Andy is funny, but one day he is going to move on to work on Linux drivers for Nvidia."

If NVIDIA won't release all the info for FOSS, . .

[Anonymous Coward]

If NVIDIA won't release all the info for FOSS, then they should release bug-free drivers. To follow their present course without any better explanation than they give is muddle-headed thinking on their part.

There are other OSs besides Linux

[dammy]

Because there are a bunch of smaller OSs out there that are getting the short end of closed source drivers. They are not big enough for the nVidia to notice for 3D support (how long ago was it nVidia didn't care about Linux?). AMD/ATI seems to be getting it's about making their hardware attractive to end users. After all, sales are sales, they don't have to care what OS is accessing their hardware.

Perhaps if they made cards with superdupersecret stuff on a ROM, charge an extra $25 and release the SDK for OSS that would meet their legal/technical requirements?

Re:There are other OSs besides Linux

[TheRaven64]

NetBSD committed DRI/DRM code a couple of weeks ago. Now, NetBSD have full support for older Radeons and Intel chips, and will get the other drivers when someone gets around to porting the (small) kernel-specific portion of the code over. I believe Haiku has also taken some of the DRI code to get 3D support.

Of course, none of these systems get 3D support with nVidia chips.

NDAs and Patents Suck Life.

[twitter]

What you say is confusing and has the smell of a well crafted lie. Can you set me straight so I can understand why Nvidia is unable to do like Intel and fully co-operate with the free software community?

Given that they make their money off of licensing IP and perhaps selling software (as opposed to selling hardware), this is highly unlikely.

That term, "IP", has confused me. I thought Nvidia made GPUs. What would they lose by telling free software developers how to make their GPUs work?

they own all of the IP for the 2d driver, some parts of the 3d support are licensed from other companies and individuals (who hold the patents). The terms of those licenses forbid nVidia from releasing their code (naturally).

This makes a little more sense, despite that awefull IP term. You are saying it's not Nvida's fault but I'm not sure how that can be true. Nvidia is huge and should be able to set terms better than that. Further, they are releasing binaries, so I'm not sure what Uncle No-No is protecting. Finally, while the free software community may be able to provide some help in convincing Uncle No-No to co-operate, this is something Nvidia could do better.

Nvidia's hostility to free software only hurts their hardware sales. Gamers crave the clean world of free software and will spend more money on cutting edge hardware if their wallets are not bled white by M$. As things stand, upgrading hardware requires a new M$ OS purchase and begging or a cracked version of M$ OS. People like me won't even consider buying a cheap or middle of the line Nvidia card because it won't do much more than heat up the room. The non free drivers are better than that, but they make my system brittle like a M$ system - flaky and harder to upgrade. A system like that is disposable and I might as well buy a Play Station. Nouveau has promise, but the vast majority of GNU/Linux users are going to wait until it's in Debian Testing. You would think they'd want to grow what's going to be a richer market for them.

Re:NDAs and Patents Suck Life.

[Dragonslicer]

Given that they make their money off of licensing IP and perhaps selling software (as opposed to selling hardware), this is highly unlikely.

That term, "IP", has confused me. I thought Nvidia made GPUs. What would they lose by telling free software developers how to make their GPUs work?

I think you misread that sentence. The "they" in that sentence is the companies that nVidia has licensed code from, not nVidia itself.

Re:NDAs and Patents Suck Life.

[twitter]

I think you misread that sentence. The "they" in that sentence is the companies that nVidia has licensed code from, not nVidia itself.

Ah, so what does that have to do with nvidia again? As I said, you would think nvidia's in charge.

Re:NDAs and Patents Suck Life.

[mikael]

What you say is confusing and has the smell of a well crafted lie. Can you set me straight so I can understand why Nvidia is unable to do like Intel and fully co-operate with the free software community?

Have a look at NVidia's OpenGL specifications web-page

Every extensions comes with an IP Status field. For example ARB_color_buffer_float has the following:

IP Status

        SGI owns US Patent #6,650,327, issued November 18, 2003. SGI
        believes this patent contains necessary IP for graphics systems
        implementing floating point (FP) rasterization and FP framebuffer
        capabilities.

        SGI will not grant the ARB royalty-free use of this IP for use in
        OpenGL, but will discuss licensing on RAND terms, on an individual
        basis with companies wishing to use this IP in the context of
        conformant OpenGL implementations. SGI does not plan to make any
        special exemption for open source implementations.

        Contact Doug Crisman at SGI Legal for the complete IP disclosure.

Re:NDAs and Patents Suck Life.

[Dragonslicer]

Ah, so what does that have to do with nvidia again? As I said, you would think nvidia's in charge.

I would guess that means that nVidia can't release the source code for the drivers because they don't actually own the rights to the code. As for nVidia being in charge, I would think that nVidia, being the far larger company that the smaller company may not be able to exist without, would have the advantage in negotiations. Then again, if nVidia said "Screw you, we'll write our own code", they'd probably get slapped by a patent infringement lawsuit. The only outcome that could possibly lead to a win for Linux users would be nVidia buying out all of the companies that they've licensed code from.

Re:NDAs and Patents Suck Life.

[jandrese]

Because nVidia uses some of that code in their driver. They can't release that code because it isn't theirs, but without it the driver is useless. I'm guessing if they could write it themselves cost effectively they would have.

It's the same reason DooM's source code wasn't released as fast as the rest of ID's projects. DooM used a licensed sound module that they had to eventually rip out of the code before they could release it.

A good example of functionality that can trip up a driver release like this is Macrovision support on TV-Out ports. In some cases they are effectively required to support (thank you DMCA), but the licensing also requires that all drivers that enable the TV Out ports must enable it, yet the code to enable it is held by companies like Macrovision who would never release it open source. Even if it's just hardware register settings they'll call it proprietary information and forbid the release.

Re:NDAs and Patents Suck Life.

[citizenr]

except that software patents work ONLY in US.

Re:NDAs and Patents Suck Life.

[Macthorpe]

Reading your posts today has been a joy. This piece of abject stupidity regarding patent licensing, and your other comments regarding GSM phones, have only proved that towing the /. party line leads directly to mod points, no matter how uninformed you actually are.

Re:NDAs and Patents Suck Life.

[PitaBred]

It's not even code licensing. It's algorithm licensing, stuff like S3TC. They could re-write it easily, but it'd still be contrary to the patents that S3 has on it.

Re:NDAs and Patents Suck Life.

[dedazo]

It doesn't really matter, does it? All your questions are rethorical, disingenuous and are predicated on the idea that intellectual property should not exist and in order to "co-operate" NVidia should release all of their source code.

I guess that and the "M$" drivel (were we talking about Microsoft?) pretty much define your position, so why should the OP even bother to "set you straight" at all?

Re:NDAs and Patents Suck Life.

[LingNoi]

Lucky for me I live in a country that doesn't believe in software patents.

This patent BS only concerns Americans and is a weak argument of not releasing anything.

Re:NDAs and Patents Suck Life.

[mr_mischief]

...and where do you think the majority of NVidia's market is?

Re:NDAs and Patents Suck Life.

[LingNoi]

The whole world maybe? .. Sorry to break this to you but US != world and US Law != World Law

Re:NDAs and Patents Suck Life.

[mikael]

These are more hardware patents than software patents, since these are implemented on the GPU of the graphics board. But the problem is that in many cases, a specific patent may apply simply in the way the source and destination memory buffers of a rendering operation are set up (texture mapping, render-to-texture). These may be as simple as setting up a pointer and some format/dimension information within the driver. Allowing programmers to write their own setup code would then have to apply for patent licenses directly.

The secret NVidia code

[maxwell demon]

What secrets is NVidia really trying to protect?

Well, here's the code NVidia tries to hide. You don't have it from me, OK?

void voodoo()
{
  user_soul* soul = get_soul_from_user();
  det_list_add(soul->determinations, create_determination(det_buy, NVidia->graphics_card, _always));
}

Re:The secret NVidia code

[Enderandrew]

Small correction:

void voodoo()
{
    user_soul* soul = get_soul_from_user();
    det_list_add(soul->determinations, create_determination(det_buy, NVidia->graphics_card, _always));
    ???
    profit++
}

Re:The secret NVidia code

[TheRaven64]

Let me guess, that was code they acquired when they bought up 3Dfx's assets?

WHO ARE "THEY"???

[Anonymous Coward]

Or is that under NDA???

Jeesus, it's like getting blood from a stone here.

Last I tried the 2d "nv" driver...

[mi]

It could not even support a dual-head setup. If that's what he means by "NVIDIA engineers actively maintain and improve", then it is simply sabotage:

• Why don't we fix the dual-head setup, boss?
• Because we want people to use "nvidia" driver instead.
• What do we care — we give it away for free anyway?
• Because we drop support for older cards from our official driver, thus pushing people to upgrade them.

The nv-driver was my only option on FreeBSD/amd64, yet it would not drive the second monitor, so I changed the card for a Radeon. The open-source driver for ATI, at least, supports dual head and plenty of other features found on the hardware.

Re:Last I tried the 2d "nv" driver...

[etrusco]

Also, the "nv" driver is clearly _much_ slower than the proprietary one (on 2D, of course).
I say TFA's argument "NVIDIA actively maintains the open source 'nv' driver" is utter bul- err, completely bogus.

But not in the 8xxx series cards

[gzunk]

Except the XvMC extension is NOT supported in the 8xxx series cards. Can't be bothered to dig out the link, but it's true. You buy a top of the line NVidia card, expecting to get at least as much as the previous card generation, but nope. Apparently it's too difficult to implement on the new architecture.

OK, that Sucks Life.

[twitter]

Thanks for the link, it clears up a lot. The site itself is a pain, so it would probably be easier to Google site:developer.nvidia.com for "IP Status".

As for the example you give, Holy Shit!

SGI owns US Patent #6,650,327, issued November 18, 2003. SGI believes this patent contains necessary IP for graphics systems implementing floating point (FP) rasterization and FP framebuffer capabilities.

A patent on floating point raterization and framebuffers? Is that what I think it is? Yes it is. I can not think of anything more obvious in high quality imaging than representing the image as a floating point matrix. It may be true that there are still "fat line" patents out there.

Kudos to Nvidia for shining a small light on this insanity. Knowing the problem is always the first step. It would be nicer if they would put patent and other encumbering as symbos on the reference page and a link to the actual patent in the description page.

Re:OK, that Sucks Life.

[dbIII]

It just shows that software patents have serious effects even outside of the USA where nobody accepts their existance.

hi twitter

[Anonymous Coward]

Just wondering if you were planning on replying to some of these. Thanks.

Re:hi twitter

[twitter]

Just wondering if you were planning on replying to some of these. Thanks.

No.

Re:hi twitter

[Anonymous Coward]

No? "No" as in "no, because I'm too embarrassed to" or "No" as in "no, because I don't really understand what the problem is"?

Re:hi twitter

[twitter]

No, as in there's nothing I have to add.

Re:hi twitter

[Anonymous Coward]

Or rather, as in you saw your daily opening to eject your "M$" bullshit and only managed to prove that you are indeed as stupid as you let on.

With friends like these, open source hardly needs Microsoft's threats to be worried.

It's bloody obvious what's really happening here..

[advocate_one]

Microsoft are the real owners of the SGI OpenGL patents and are blocking this entire show by insisting on those anti-GPL RAND terms for the licensing... not SGI...

http://www.theregister.co.uk/2002/01/16/sgi_transf ers_3d_graphics_patents/

http://www.smithhopen.com/news_briefs_display.asp? ID=301

http://www.forrester.com/Research/LegacyIT/Excerpt /0,7208,28681,00.html

that last is is a doozy... they want $99 for a one page article...

Microsoft has nVidia over the certification barrel... if they make the nv driver support 3D, then nVidia may find it very difficult to get their windows drivers certified... they're having enough problems at the moment...

SGI aren't the problem

[Anonymous Coward]

As far as SGI have said (very publicly) there is nothing that SGI have patents for that they don't have a problem being GPL'd

Re:SGI aren't the problem

[ichigo 2.0]

They could always change their mind. The GPLv3 would help in this case only if SGI themselves distributed/contributed code that covered the patent in question. If they are really serious about not minding GPL code that use those patents, maybe they will release something under v3 themselves? Just saying something publicly doesn't make it legally binding.

Thanks, that was not obvious.

[twitter]

Microsoft are the real owners of the SGI OpenGL patents and are blocking this entire show ...

Ah yes, who else can suck life like M$? I should have known.

Re:Thanks, that was not obvious.

[Anonymous Coward]

I'm thinking that "life" may not be the most fitting word there...

Excuses do not a tech support market make

[WillAffleckUW]

And, to be honest, more support is needed than is currently provided.

Hey now that's nasty

[BillGatesLoveChild]

> I'm sure they would fix it

Well, you'd think that, wouldn't you? So why then are there so many web pages talking about it?

>> Especially when the first post you linked to says
> so i got the 91.28 drivers and they finally worked, now i can play games like i did before

I've upgraded the drivers so many times and tried all those tweaks. I still get it. Today was a bad day. Sometimes it'll go for week without crashes. Imagine my delight at lost work, yesterday a corrupted word file.

> If you're claiming it's never been fixed yet there's step-by-step solutions in all of
> the "evidence" you linked to, you either cannot read (unlikely, you can write) or you are trolling.

Why would I make something like that up? Hell man: The Windows BSOD Error message says "an infinite loop has been detected in the nv4_disp.dll driver and is being shut down. This is probably due to a device driver programming error". I spend too much time rebooting and *you* call me a troll? Can I call you an nVidia astroturfer? (It's cheaper than providing tech support.)

Flamebait my ass

[BillGatesLoveChild]

Let's see: nVidia don't release their driver source, don't fix their driver's bugs, the infinite loop bug is widely reported, yet, somehow, reporting this is flamebait?

As Slashdot tells you: "use your mod points wisely"

Mini-RMSes

[Hucko]

I'm just wondering why so many open source users have disdain for companies not open sourcing their software when it is potentially against their aims to do so. I mean, to the exclusion of actually using the software which could make their computer experience better. Surely we haven't got that many mini-RMSes?

I'd be a pseudo-RMS. If it made the experience better then yes, it has been (in the past) compulsory to maybe consider using the software. Sometimes the initial experience was great! But a minor upgrade / alternative program / (pick a variable) would break it and I'd have to start again.

Why do you _expect_ a company to do business according to your own rules?

Because I believe it is just as unreasonable of a 'non-entity'* to trust them, just because I'm paying for it. I want to be able to see a reason to base that trust on.

* recognising that they are people too, but the responsible persons can change at the drop of a hat. I've seen management at work. All they do is prevent good ideas becoming successful for bureaucratic reasons.

Grass is always Greener (Come back Intel!)

[BillGatesLoveChild]

> The other thing is, when I have experienced problems with the nVidia drives and reported them,
> I have always been contacted back by nVidia in an attempt to resolve the problem.

This is for the nVidia Linux driver? I haven't been able to find anyone who got a response from nVidia about anything, let alone the infinite loop bug. The problem doesn't have to be the bug per se, but the fact they won't fix it or even acknowledge its existence.

> I curse the drivers for my ATI card daily at work.

Ouch. As someone in the nVidia paddock, the grass isn't any better here!

Well Intel have announced they're re-entering the 3D Graphics field. Their original 3D card was a rock-solid effort: It never crashed, not once, and they provided drivers years after it had been withdrawn from market. Hopefully Intel will come and give nVidia and ATI the butt kickings they so richly deserve. Welcome back, Intel!!!

Re:Grass is always Greener (Come back Intel!)

[AaronW]

I run both. I have rarely had any problems with my nVidia 6200 and 6600 and have never experienced the problem you describe. I have used both systems for gaming under Linux and more. I used nVidia's bug reporting mechanism when I did have problems and they did respond (though were not able to help - backporting the driver worked and a newer driver fixed the bug). As for Intel, my desktop at work has an integrated Intel chip which doesn't seem to work properly with Xorg 7.2 (nor can it drive two monitors).

Re:Grass is always Greener (Come back Intel!)

[BillGatesLoveChild]

Where is nVidia's bug reporting mechanism BTW? I've never been able to find one. Their website says if you purchased your product directly from nVidia, fine, but otherwise talk to your OEM. They used to have a feedback form page that was always 'under construction'. They've removed that entirely now (although just noticed added a new one for developers): http://www.nvidia.com/page/support.html

The infinite loop bug is hellacious, but either you've got it or you don't.

> As for Intel, my desktop at work has an integrated Intel chip which doesn't seem to work properly with Xorg 7.2
Intel's integrated i9xx chipset has 3D built-in. I've got it on my lappie. It's not a bad implementation, but definitely lowest common denominator and won't play newer games. When they announced their re-entry to the 3D field, they got a lot of flack for this chipset. I'm hoping Intel's 3D team will feel the heat from that. (I'm naive, but occasionally good things happen . :-)

Hypothetical Scenario that hasn't happened

[Prien715]

In the future, when new versions and extensions of OpenGL are released, we won't have any guarantee that they will properly update the drivers.

NVidia has been releasing binary Linux drivers for over 5 years. There's been new versions of OpenGL, new extensions, the new NVSG, bugs, and all this time NVidia has done a good job releasing drivers that are up-to-date with their Windows counterparts and stable (can you say the same for ATI?) In my 8+ years using Linux, I have never had a problem with an NVidia driver (except one where I submitted a bug report and it got fixed). The hypothetical scenario you mention has, to my knowledge, not yet happened even though there was a much better business case for having poor Linux support years ago than today.

Nowadays, the choice is clear: go Intel X3000/X3500, which supports open source and you can be sure will always be up-to-date

By up-to-date you mean...using a video card that performs worse than 3-year old NVidia hardware? People have different priorities. For me, if I buy a 3D card, I want performance (for 2D, there's no argument since NV's 2D driver is open source and outperforms intel). I don't want to buy something that's already out of date.

That's sure some commitment...

[Chris Snook]

If Nvidia is committed to making the 2D nv driver work, then why the @#$% doesn't dual-monitor support work? It's not that hard, it just requires knowing the hardware. There's not a shred of OpenGL involved.