Mixing open and closed source

[Redeeman]

Intel seems to be leaving out the same features out of their docs, and NVidia isn't releasing any docs at all. Tell me again how this makes us the bad guys ?

i never said you were the only bad guys. Also, i think you might have misunderstood what i said with "bad guys" - i meant that between you and the OEMS(which i have nothing to do with), from my perspective, if i buy an ati card, its your "fault" some features are unavailable due to DRM. Even though the OEMs might pressure you

[givemesugarr]

That is correct... But again directly swapping nouns around... Quoting you directly with swapped nouns...

for what i know DRM ensures that no copies of the media are made and sold to someone without paying the appropriate fee and that you only make copies of the music under the direct guidance of the copyright owner. and this is to protect the artists rights. i ask for forgiveness if i misunderstood this.

Note that I am not arguing the morality (for profit vs for freedom), but rather that DRM is a technical means of enforcing the media companies view of fair use, and is fundamentally no different than the kernel authors enforcing their view of fair use.

Please don't interpret this as a flameworth response, I am just trying to ensure that you have a clearer understanding of my point.

Regards.

the difference is:
you know what the kernel does and how it does it, you're not forced to pay anything and you can use it in the way you like and distribute it freely. the drm block users from using features like the hw hd video decoding for which you've paid in the price of your hw board that you legally bought. so the 2 things may at look like the 2 sides of the asme medal, but they're not. drm deprives people from their rights while gpl ensures that people have always their rights. so they're 2 different things.

[givemesugarr]

Although you seemed to have implied that my comments were about the GPL license and not the GPL_ONLY flag in the kernel...

DRM is technical implementation to support the same thing. It is uses copyright laws (including DMCA) to ensure that no unauthorised copies are made no matter what (the no matter what may be naive, but it is still there).

what about a legally bought stuff that you want to copy for personal backup (this is a legal practice in many world countries)?!

Again, ignoring the morality nor the intent (for profit vs for freedom), DRM and the GPL_ONLY flag in the kernel is fundamentally the same thing.

again: they're the opposite! one deprives people from freedom, while the other ensures the freedom of the users.

I made no comparison between the GPL and DRM, I made a comparsion between the GPL_ONLY flag in the kernel and DRM as both being Technical measures used by a copyright owner to enforce their interpretation of fair use.

You obviously didn't click the link that was quoted, since both Linus Torvalds and Theodore Tso have both made the exact same comparison.

Quoting from the thread within the link.

depriving people from some rights while ensuring the people they do have the rights is not the same thing as you seem to say. also, if we look at the real topic of this thread: drm deprives users from hw functionalities for which they've paid just because someone wants this to happen and to favour some other individuals.

Let me re-iterate one more time. DRM is not fundamentally any more evil than the GPL_ONLY flag. Users get stung by DRM in the same way that naive changes to the kernel developers sting non-GPL module developers.

i don't know of any function due to gpl_only that didn't let the user to exercise his normal freedom and rights, while i know a lot of this with drm. just think of the old legal bought drm music from internet shops that are usable just with the old readers that were present when you bought it. think that you cannot listen it if not with that old reader and that when you change it, if you still want to listen to that music you'll either have to rebuy it or to use the old reader. i don't know of something similar to have happened due to gpl_only flag.

For example, GPL_ONLY is intended to support a developers view that their interface is Linux unique, and any modules using that interface is clearly a derivation of the kernel, and consequently must be GPL.

In no way, should a non-GPL module be considered a deriviative work if it uses a standard Unix function such as udelay. But yet, if PARAVIRT is enabled in the kernel, the udelay function gets hidden behind the GPL_ONLY data structure and the compile fails.

As Linus and Theodore indicate, any form of fair use enforcement is fundamentally broken in the same way. Evil is purely a point of view.

well, i don't know of this issue and if this would be real i think that xen wouldn't be usable, but it is so i don't think this is a real issue.

[givemesugarr]

I disagree. No distribution would remove the EXPORT_SYMBOL_GPL flag from the functions. So it DOES prevent people from doing things in the real world. Asking for people who operate in the media market (like AMD) to ignore DRM is fundamentally no different than asking distribution to strip off the GPL_ONLY flags from the kernel. If you part of the ecosystem you must accept the technical measures of that ecosystem. For the media playback industry, DRM is a requirement.

wrong. you could switch distro for some that it uses it, or you can chose gentoo on which you could remove it manually.
drm is a hw stuff that is mandatory to have on hw for them to be compliant to some hw certification. and i don't see what the real problem is since you know what that flag does and how it does it, which is not true with drm.

As I have said in another post. For 2.6.22, if you enable PARAVIRT in the kernel suddenly udelay - a function that exists in nearly any unix style system in the universe, is reported as a GPL_ONLY function.

and what's the problem with that?! does this prevents you from using the function in the way you like?! i don't think so. drm instead prevents you from doing something the way you like.

Just like DRM, it is a fundamentally broken way

In the same way that DRM [CSS, AACS, etc] (ultimately - invariably it gets cracked) does stop people from using their FAIR USE rights.

I am always amazed at the number of people who openly flout DRM, but then cry foul about people removing the GPL_ONLY flag. Both are removing technical measures that are intended to guide against copyright violations.

what does css have to do with this?! since when the style sheets prevent you from doing something?!
the aacs is the thing that we are really discussing here, since it is hw implemented in a portion of hw that could be used for accelerated free video decoding and not only for protected one. this portion, though cannot be documented to the public since it contains the drm stuff in its module and this could break the module. so in this case, drm prevents the user from having a functionality that it has the rights to have but others prevents him from doing it. i've never heard for something similar happening with the export_symbol_gpl as you've argued.
again, i don't think that you can compare hw things to sw ones. the drm problem is hw while the export_gpl is sw. i don't know how you can compare 2 different things between them.
also, you cannot compare something that ensures that you always can control something and have the right to look at how it does things and something that you cannot control and that prevents you from doing legal stuff like backup copies.

[bridgman]

what does css have to do with this?! since when the style sheets prevent you from doing something?!

CSS is (like DRM) one of those unfortunate acronyms which means many things. CSS is also the encryption algorithm used on regular DVDs -- what you remove with deCSS.

I have to admit that cascading style sheets confuse me and stop me from doing what I want much more than DRM

[Michael]

what does css have to do with this?! since when the style sheets prevent you from doing something?!

LOL, good one

[Svartalf]

I have to admit that cascading style sheets confuse me and stop me from doing what I want much more than DRM

Nice rejoinder, John...

[GDJacobs]

I don't care about the ability to decode DRM'd crap.

I just want decent hardware video decoding for un-DRM'd stuff.

No offense, I don't know 1 person who owns something outside of an xbox that has an optical device for HD playback. Not one. Yet I still know tons of people who watch various unencrypted h264, xVid, theora, mpg4, etc., on their computers. The only, and I mean only, DRM'ed format I see people using is DVD, which everyone and their brother circumvents anyhow, whether it's because they want to make backups or skip the warnings, ads, and trailers.

You guys are just part of a collusive anti-consumer industry, which must suck.

Personally, instead of modularity, I think you should just do one open source driver that works right instead of having all this monkey business with like 3 drivers, none of which actually work well.

I do appreciate your guys efforts, and I know hearing this stuff must be frustrating and discouraging, but there are those of us who own ati cards which no matter what driver we use basically function at like 60% of what they should because with every driver there is some stupid trade-off based around stuff like this discussion.

Just give me working 3-D, decent video playback for unprotected video, and monitor autoconfiguration that actually works properly in one driver and I'll be happy. Until you've done that, worrying about DRM is putting the cart before the horse.

Really? I want GPGPU support. Once we have a general way of doing shader math, projects like ffmpeg, SETI@Home as well as pretty much everyone into scientific computing will be able to at least have the possibility of offloading computation onto the graphics card. This is something I suspect AMD is interested in, with Fusion just around the corner.

[bitnick]

Sorry for the long post, but just like everybody else I guess I feel a need to ventilate my feelings on the subject, and I also wanted to explain to bridgman (and any other AMD/ATI employee) how I feel and why, so that they don't feel too depressed about the way they might be treated.


Of course AMD/ATI are not without fault when it comes to DRM.

Of course they are not the only GPU producer that's not without fault. This doesn't in any way excuse a company's acceptance of DRM, though (although it does excuse the inclusion of it in their products). Mixing DRM and decoding engines so that it is not possible to use non-DRM related things (like decoding) without exposing the DRM clearly shows a deep acceptance of DRM - otherwise the sane thing would be to cleanly separate the two.

Of course there's a lot of pent up anger and frustration with regards to DRM, which is going to spew out through any available vent. However, even if talking about DRM means the company will have to take a lot of shit (justified or not), I still feel that showing that they are ready to talk about it and explain their view gives more good-will than bad-will.

Of course there's also a lot of pent up anger and frustration among Linux users with regards to the crappy support of ATI cards. I have a Radeon 9500, bought long ago while I still used Windows, and a Radeon X800 that I got from a friend. Despite the countless hours I've spent trying to get these to work, both cards are still seriously crippled. So far, the documentation released and the work done has not changed my experience with my cards.
Even if there was a new driver released tomorrow which solved all problems, that would not compensate me for all my hours of work and the years of not being able to fully use what I payed for. ATI as a company should not expect to be thanked for finally releasing what their customers have been paying for all along; the best they can hope for is forgiveness for not doing so sooner. And that will only happen when all the documentation is out and we get drivers that actually work, i.e. after "fait accompli". Until then ATI (and thus unfortunately their employees) will have to expect being treated with a huge amount of scepsis. Sure, ATI is now AMD/ATI, and the open source people might be newly hired, but we are still talking about the same ATI graphic cards. And most importantly the drivers for my cards are still the same as before.


If we do get truly working 2D, 3D, IDCT and MC acceleration for ATI cards in an open source driver (and not just for a few GPU families, but ongoing support), I think this would be a great thing for the Linux community! Especially along with the same support for Intel chips. It would suddenly become meaningful to write games for Linux, and a whole new category of users (gamers) would start to look at Linux as something usable. It would mean that ordinary desktop users could get "cool" 3D desktops and watch multimedia (albeit only non-DRM-damaged material) without having to become IT experts on the way. The fact that DRM:ed material would not work while other material would work perfectly would raise awareness of what DRM is among "ordinary people".

I'm sure the content industry and Microsoft are aware of the importance of functioning graphics drivers on the Linux platform, and I'm sure they are doing and will continue to do everything in their power to make the realisation of these drivers difficult (the true reason for Microsoft's wholehearted support of DRM, perhaps?). I don't expect AMD/ATI to make a stand for OSS drivers and making sure we get them in spite of the demands from the content industry and MS just to be nice; however OSS drivers could be a way/excuse for AMD/ATI to keep control over their own products. As a bonus (for AMD/ATI) and plaster (for the Linux community) it could also be a nice way for AMD/ATI to make up for ATI's past shortcomings. So OSS drivers could save both our days.


Yes, I *am* angry with ATI, as I am with every other entity that has been just sitting there swallowing this DRM crap and/or ignoring the open source community, mumbling about intellectual property (damn it - if you presume to sell hardware, then you *will* release the documentation necessary for the buyers to use it!), but I am also happy that there might now finally be documentation and drivers on the way!

[mgc8]

That is correct... But again directly swapping nouns around... Quoting you directly with swapped nouns...
Note that I am not arguing the morality (for profit vs for freedom)

Your argument is completely flawed due to this very point: you can not separate the morality from the action. Doing that would throw us into a relativistic soup where I could for example say the following (keeping within your style):

1. "I want to defend my rights, among them the right to live, therefore I will enforce them even resorting to violence against someone who threatens me"
2. "I want to defend my rights, among them the right to kill people and eat them, therefore I will enforce them even resorting to violence against someone who threatens me".

If you "ignore the morality", then the two phrases above are totally equivalent, even though obviously antagonic. There really is no way to decide which would be "right". However, would you agree with number 2 ?

Of course GPL uses the same technical means that DRM uses, as the very idea of GPL was to turn copyright on it's head -- but it's all about the morality in this discussion, and taking DRM's side is like saying that police officers are "infringing the killers rights to maim and murder".

for what i know DRM ensures that no copies of the media are made and sold to someone without paying the appropriate fee and that you only make copies of the music under the direct guidance of the copyright owner. and this is to protect the artists rights. i ask for forgiveness if i misunderstood this.

There is a big difference between "copyright owner" and "artists". Almost all artists sign contracts that give away their "rights" to someone else. Can you guess who that someone else is?