From francisco@imagine-sw.com Thu, 03 Feb 2000 10:23:00 -0500 Date: Thu, 03 Feb 2000 10:23:00 -0500 From: Francisco Figueirido francisco@imagine-sw.com Subject: [Livid-dev] How interrelated is CSS with MPEG2 decoding? This is a question/comment that might strike some of you as rather naive (I was going to write `stupid' but somebody once said that there are no stupid questions :-). I confess that I am still rather uninformed, although I have been following this list for a couple of weeks. My question/comment was spurred by what a person related to Sigma Designs said to me yesterday at the LinuxWorld expo. I asked him (the obvious question, I guess) when, if ever, could we expect support for the DXR3-based cards (like Creative Labs' Encore). He said, quite plainly: NEVER. According to him, it won't ever happen because they cannot provide a software implementation of CSS decryption (even if binary) without violating their license. I ask him whether they could consider providing a driver without any CSS-related code but enough hooks but he considered that approach hypocritical (no flames, please!). I am inclined to agree with him, although I fully believe the DVD Forum is way out of scale on any morality issue. Anyway, only later I realized that there is no logical need for a DVD (movie, I mean) to be CSS-encrypted. If I understand correctly, only followers of the DVD Forum will provide CSS-encrypted movies. For example, when I bought my decoder card (Creative Labs) I got a DVD from National Geographic, which is NOT encrypted. And I can imagine many other small producers (in fact, I have a friend who is thinking about producing teaching materials in this way) will do the same. What this means, to me, is that the card manufacturer should be LEGALLY able to provide drivers to read unencrypted MPEG2 files (in fact, the person from Sigma Designs asked me whether all I wanted was to play MPEG2 files; I should have answered: YES). Of course, it would not be unnoticed that once one can send an MPEG2 file to the decoder, reading an encrypted movie is a matter of (illegally?) decoding it first ... Unless I am missing something, which is quite likely. Thanks for your time/bandwith. -- Francisco Figueirido, Ph.D. Phone: (212)317-7680 Quantitative Analyst Fax: (212)317-7601 Imagine Software, Inc. e-mail: francisco@imagine-sw.com 400 Madison Avenue, 21st Floor New York, NY 10017
From jfreed@johnfreed.com Sat, 5 Feb 2000 13:07:11 -0500 (EST) Date: Sat, 5 Feb 2000 13:07:11 -0500 (EST) From: John Freed jfreed@johnfreed.com Subject: [Livid-dev] How interrelated is CSS with MPEG2 decoding? On Thu, 3 Feb 2000, Francisco Figueirido wrote: > My question/comment was spurred by what a person related to Sigma Designs > said to me yesterday at the LinuxWorld expo. I asked him (the obvious > question, I guess) when, if ever, could we expect support for the DXR3-based > cards (like Creative Labs' Encore). He said, quite plainly: NEVER. According > to him, it won't ever happen because they cannot provide a software > implementation of CSS decryption (even if binary) without violating their > license. I ask him whether they could consider providing a driver without > any CSS-related code but enough hooks but he considered that approach > hypocritical (no flames, please!). Although this is bad for you, it might be a Good Thing for the lawsuit, as it demonstrates that vendors are scared to provide *any* sort of Linux interoperability, even when it would not conflict with CSS. Further, and this suprised me, the DVD Forum has sworn in court that it *would* grant a license (did you see this, Alan Cox?) to any Linux developer willing to pay the $10,000 and sign the NDA. He also asserted that the problem is not the making of copies, but the making of *perfect* copies. They don't object to analog copies because of the inherent degrading of the signal across multiple generations. http://www.eff.org/ip/Video/DVDCCA_case/20000114-pi-hoy-rep-dec/ 20000114-pi-hoy-rep-dec-11.gif "If a person or entity were prepared to take a license on the same terms as existing licensees, such a license would be granted." Note that he says "would" not "might". John Freed
From pvolcko@concentric.net Sat, 5 Feb 2000 13:44:04 -0500 (EST) Date: Sat, 5 Feb 2000 13:44:04 -0500 (EST) From: pvolcko@concentric.net pvolcko@concentric.net Subject: [Livid-dev] How interrelated is CSS with MPEG2 decoding? > > My question/comment was spurred by what a person related to Sigma Designs > > said to me yesterday at the LinuxWorld expo. I asked him (the obvious > > question, I guess) when, if ever, could we expect support for the DXR3-based > > cards (like Creative Labs' Encore). He said, quite plainly: NEVER. According > > to him, it won't ever happen because they cannot provide a software > > implementation of CSS decryption (even if binary) without violating their > > license. I ask him whether they could consider providing a driver without > > any CSS-related code but enough hooks but he considered that approach > > hypocritical (no flames, please!). > > Although this is bad for you, it might be a Good Thing for the lawsuit, > as it demonstrates that vendors are scared to provide *any* sort of Linux > interoperability, even when it would not conflict with CSS. Well, put simply you nor anyone else free to speak about it, knows the terms of the CSS NDA/License. Without judging the merits or intent of the license, it is very reasonable to assume that the license prohibits the "hooks" that you suggest. The same logic applies to why they can not release the driver for the older generation cards, it would force them to break terms of their licensing. And the license such that it is, doesn't allow them to speak about the terms of the license or go into much (if any ) detail as to hypothetical implementations. I totally understand why a lot of people are pissed off about this. It doesn't make much sense logically and or morally/ethically. But the legalities of contracts never really seem to deal with those paticular aspects of life. I own a hollywood+ card as many of you do. I'm displeased with the inability (or lack of motivation, however you want to view it) on Sigma's part to support this card under linux. However, it is also clear that Sigma has at the very least made up it's collecitve mind on this issue and they seem to have significant legal reason for that judgement. As such we should really be supporting Sigma here. They are the first and only vendor to come out and officially write drivers/support for the linux operating system. Matrox and Creative may have iven some docs, but they aren't actively putting their own dollars into drivers. In fact they are offering very little in terms of any support at all for those trudging through the specs and code, from what I understand. This level of support from Sigma is important. And although their first "beta" level drivers will likely be far from what we are used to, I'm confident that Sigma will put the time and effort into continued support and development of the drivers and user space libs so that they come to the point were we are happy as linux users and developers and they are happy as well. Let's applaud Sigma designs and show them support for their efforts. Perhaps it isn't perfect, maybe even far from perfect, but at least their putting their money where their mouth is and actually getting it done. Thats more than can be said for Creative labs (DXR2 support) or Matrox (G200/400 support). Through support (and constructive criticism) I think we'll effect the final product that Sigma Designs releases. Through nit picking and bitching all we'll do it turn their decision makers away from linux in general. > Further, and this suprised me, the DVD Forum has sworn in court that it > *would* grant a license (did you see this, Alan Cox?) to any Linux > developer willing to pay the $10,000 and sign the NDA. He also asserted > that the problem is not the making of copies, but the making of *perfect* > copies. They don't object to analog copies because of the inherent > degrading of the signal across multiple generations. > > http://www.eff.org/ip/Video/DVDCCA_case/20000114-pi-hoy-rep-dec/ > 20000114-pi-hoy-rep-dec-11.gif > > "If a person or entity were prepared to take a license on the same terms > as existing licensees, such a license would be granted." > > Note that he says "would" not "might". Of course they would. Problem is that most companies and people involved in linux development won't accept the same terms that other vendors have accepted because it goes against their philosphy of free source code and open development. Not to mention there is serious question as to if the terms can even be met due to the openness of the kernel source code on which the DVD related app would be running. The problem here isn't coming up with money. It is having those with the money feel good about the terms of the licensing and what the app would have to be in order to meet those restrictions. LSDVD has approached several companies within the linux world and have been turned down not because of the lack of market, or a bad proposal, or anything like that. It has been because the companies will not support or buy into an app that must be (at least in significant part) closed source. If there is a hardware or software compnay out there who would be willing to do this, we haven't met them yet. Paul Volcko LSDVD
From ekl@wans.net Mon, 07 Feb 2000 15:28:57 -0600 Date: Mon, 07 Feb 2000 15:28:57 -0600 From: Evan Langlois ekl@wans.net Subject: [Livid-dev] CSS hooks > Further, and this suprised me, the DVD Forum has sworn in court that it > *would* grant a license (did you see this, Alan Cox?) to any Linux > developer willing to pay the $10,000 and sign the NDA. He also asserted > that the problem is not the making of copies, but the making of *perfect* > copies. They don't object to analog copies because of the inherent > degrading of the signal across multiple generations. I'll just make a copy with the encryption intact. Play it back on any player - duh! As for the $10K fee. Screw them. Pay them $10K for what? > Well, put simply you nor anyone else free to speak about it, knows the terms > of the CSS NDA/License. Without judging the merits or intent of the license, > it is very reasonable to assume that the license prohibits the "hooks" that > you suggest. The same logic applies to why they can not release the driver > for the older generation cards, it would force them to break terms of their > licensing. And the license such that it is, doesn't allow them to speak about > the terms of the license or go into much (if any ) detail as to hypothetical > implementations. > We don't need "hooks" - have the driver ignore CSS completely and just play back mpeg2 streams. Assume the disk is unencrypted. > Let's applaud Sigma designs and show them support for their efforts. Perhaps > it isn't perfect, maybe even far from perfect, but at least their putting > their money where their mouth is and actually getting it done. Thats more > than can be said for Creative labs (DXR2 support) or Matrox (G200/400 > support). Through support (and constructive criticism) I think we'll effect > the final product that Sigma Designs releases. Through nit picking and > bitching all we'll do it turn their decision makers away from linux in > general. > The DXR2 works, buy one, use it. Putting down Creative in favor of a product that isn't even released does NOT help anyone. Lets not bash the few developers we got ok?