Page 12 of 17 FirstFirst ... 21011121314 ... LastLast
Results 111 to 120 of 162

Thread: ATI Evergreen 3D Code May Soon Go Into Gallium3D

  1. #111
    Join Date
    Oct 2009
    Posts
    2,081

    Default

    Quote Originally Posted by rohcQaH View Post
    uhm, what?

    When jail time is an option, you'd think he'd have different issues to contemplate than his "usefulness" to whatever cause.
    Jail time was probably a bit of an exaggeration. Maybe not. Of course you realize that you risk jail time just by waking up in the morning -- the simple chance that some dipsh** cop decides to attack you just to meet quota, or just because your skin is a different color than his. The question is HOW MUCH of a risk are you willing to take in order to accomplish something good?

    And as I said, obviously in his case, the risk exceeds the benefits -- both to him personally, as well as to society in general.

    Also remember that overseeing the OSS-drivers is only half of bridgman's job. He's also responsible for some unspecified things related to that DRM stuff he likes to talk about so often (what a coincidence ).
    So.....?
    So, should someone break their DRM, guess who'll have to pick up the pieces? I wouldn't assume he'd be too happy about that.
    If there are no pieces big enough to pick up?

    last but not least, RE'ing their DRM would change a lot. Ok, you can watch your videos GPU accelerated for a while.
    Remember that the DRM is NOT the objective here... the UVD is the objective. The DRM *MIGHT* just be a casualty of war... assuming that you can even figure out what its doing with all that encryption going on... I certainly wouldn't waste time trying to break DRM at that level -- much better idea to just continue along the same efforts as the doom9/vlc's libaacs/libbdplus work.
    But even if bridgman manages to release both AMD's and NVIDIAs DRM, it'll cost both companies money,
    Short term cost for long term gain... get that DRM into a separate DRM-block so that it doesn't break when the next version of UVD comes out. Then nobody has to worry about it. The problem is that there is a risk associated with releasing the UVD specs -- a risk that wouldn't even exist if the thing was designed with open source support in mind. So a little short term expense so that there is no longer risk of losing DRM when the UVD docs are released.
    and guess where budget cuts hit first? Maybe some OSS effort which doesn't have a measurable ROI anyway?
    I think that if something of this magnitude happened, that if DRM was broken as a side-effect of working out UVD, then a proper analysis of HOW that break happened would reveal that DRM wasn't even the objective! Which makes it DANGEROUS to put it so close to something that WAS the objective! And so it wouldn't hit against open source efforts because open source isn't even an OPTION -- it is a necessity... and it is much bigger than the 1% that the msflunkies would have you believe.
    And one way or another, the MPAA is going to react, and that rarely turns out well for their customers.
    So they'll deny new DRM-infested titles on AMD chips. Cry me a river. The customer might not like this in the short term, but long term, boycotting DRM might help.
    The root cause we'd all like to see destroyed is a totally different one. DRM is just a symptom.
    You mean government regulation? Because that is the problem where this originates. Deregulation would put the industries into proper and natural competition, which would enable ways around DRM and ultimately a dropping of DRM since it just causes too much headache for the customers when the industries can't form themselves into such unnatural cartels.

    It would work itself out like this;
    1) First stage is vendor A decides they don't want to intentionally limit their customers any more, so they drop HDCP.
    2) MPAA learns of this, so blacklists vendor A's keyset.
    3) In the face of no longer being able to provide DRM content to their customers any more, vendor A hires Hack-RM to develop a way around the DRM so that their products continue to provide the content.
    4) Vendor B sees this and decides to follow vendor A and switches to Hack-RM.
    5) MPAA tries to figure out ways to deal with this hack and implements DRMv2 and enables vendors C through F with it.
    6) A new company arrives... MPBB. They have a great new scheme for DRM, so they sell it to some of the content providers saying that MPAA's DRMv1 got broken, how can you trust their DRMv2.... so now we have a competing DRM standard.
    7) Some of the content providers switch to MPBB's DRM and start shipping stuff with it.
    8) Hack-RM starts working on MPBB's new DRM and MPAA's DRMv2.
    9) For technical reasons, vendor D and F can't upgrade any of their equipment to MPBB or DRMv2, so if any of their customers want to see any of that, they'll need new equipment.
    10) Hack-RM has gotten around MPBB.
    11) D decides to software-upgrade their old customers to Hack-RM for both MPBB and DRMv2. New hardware gets official DRMv2, Hack-RM for MPBB.
    12) F decides to forget about their old customers altogether, and go official for both DRMv2 and MPBB.
    13) Content providers are selling content in DRMv1, DRMv2, and MPBB.

    What hardware will deal with what?

    14) Customers get wind of the entire situation and realize that D's OLD hardware can decode EVERYTHING.
    15) Some hardware cloner in China starts cloning D's OLD hardware and it starts spreading.
    16) At some point as this spirals out of control, the customers are getting thoroughly confused about what device does what. New forms of DRM are showing up every day, nothing will play content in ALL of the forms. People end up buying content for a while and quickly realize that its a crapshoot whether it will play or not -- virtually ALL begin resorting to downloading the content decrypted from the internet and quit purchasing content. At least from the internet, they know that it'll play for them.
    17) facing a huge drop in sales, the content providers realize that it is because of the confusing situation with regards to DRM. They say f-it and begin selling their content DRM-free, because at least SOME of the consumers will buy it, with the confidence that it will actually WORK.

    And, of course, somewhere in the mix, MPAA tries to form a cartel among content providers. Some of them join, some of still don't like MPAA. It doesn't last long because they're all trying to undermine each other's bottom line.


    And that is how DRM would die without government regulation.

  2. #112
    Join Date
    Oct 2009
    Posts
    2,081

    Default

    Note: I'm only discussing the theoretical aspects of this. I'm not advocating that anyone go off and actually do it.

  3. #113
    Join Date
    Nov 2008
    Posts
    764

    Default

    Quote Originally Posted by droidhacker View Post
    You mean government regulation?
    nope.

    I'd disagree with some of your predictions about the outcome; in any case there are a couple of other scenarios with outcomes much worse for us. Even if I had a full UVD+DRM documentation on my desk and the means to publish it anonymously, I don't think I would. IMHO it's not worth the risk and the collateral damage. But you're free to disagree with that.

  4. #114
    Join Date
    Nov 2008
    Location
    Germany
    Posts
    5,411

    Default

    Quote Originally Posted by droidhacker View Post
    The output was already known withOUT needing the hdcp master key IN THE CASES where input data was not encrypted. I.e. non-encrypted in, no-hdcp out.

    Again, hdcp master key doesn't help with this.

    Probably the reason why the UVD hasn't been reverse engineered is because there is more interest in implementing things that don't HAVE to be reverse engineered.

    Note that nouveau hasn't reverse engineered nvidia's vdpau hardware either, even though that driver is virtually 100% based on reverse engineering.
    right. thank you to lighting me up

  5. #115
    Join Date
    Nov 2008
    Location
    Germany
    Posts
    5,411

    Default

    Quote Originally Posted by droidhacker View Post
    How would reverse engineering it "screw anyone over"? The problem isn't that they don't want UVD to be implemented in the open source drivers, the problem is that they don't have the legal go-ahead to release the required DOCUMENTATION. If the functions can be figured out WITHOUT AMD DOCUMENTATION, this would NOT be screwing anyone over.
    smart man thats the clue ....

    but yes maybe it hurts amd if the amd chips are the only one with broken video unit..

  6. #116
    Join Date
    Nov 2008
    Location
    Germany
    Posts
    5,411

    Default

    Quote Originally Posted by V!NCENT View Post
    If the _AMD_ implementation of the DRM was cracked then pirates would only laugh and jawn (ripped versions are on the web anyway) and a legeon of angry FLOSS supporters would go after the cracker with burning torches for making AMD stop giving out documentation and working drivers, me included.

    I'm serious.
    i'm sure that not Linux users are not the "warez""Moviez" people-

    Linux users try to be legal they try to make his own movies "Blender" and they try to make there own games.

    the linux users are not the problem.

    the real problem is an windows usere with no windows lizence and all games and programms are warez and all movies are ripped.

    in my point of view warez/moviez hurt linux and opensource thats because if people don't wana spend money and they can't use warez or ripps they are forced to use opensource software to save money.

    linux really not the problem.

    the real problem is the windows-only-copyprotectet-anti-linux-hardware.

    an 100% secure anti-warez hardware is 100% ok für linux users if the hardware can be used by linus users but if the hardware just hurt linux users then the linux users forced to fight for the freedome of the use of the hardware.

    in my point of view amd need to dev a shader based video acceleration to just stop hurt linux users.

  7. #117
    Join Date
    Aug 2008
    Location
    Finland
    Posts
    1,578

    Default

    Quote Originally Posted by droidhacker View Post
    So they'll deny new DRM-infested titles on AMD chips. Cry me a river. The customer might not like this in the short term, but long term, boycotting DRM might help.
    Think of the big picture. This would lead into Microsoft AMD GPU's to lose Microsoft certificates since they can't do DRM anymore. This would lead to that OEM wouldn't be allowed to sell the cheap versions of Windows with computers that have AMD GPU's. Think what this would do to sales.

  8. #118
    Join Date
    Aug 2008
    Location
    Finland
    Posts
    1,578

    Default

    Ehm, "would lead into AMD GPU's to lose Microsoft's certificates" to clarify.

  9. #119
    Join Date
    Oct 2009
    Posts
    2,081

    Default

    Quote Originally Posted by nanonyme View Post
    Think of the big picture. This would lead into Microsoft AMD GPU's to lose Microsoft certificates since they can't do DRM anymore. This would lead to that OEM wouldn't be allowed to sell the cheap versions of Windows with computers that have AMD GPU's. Think what this would do to sales.
    Where do you get that from?
    If a MAJOR GPU vendor, like AMD suddenly was TOTALLY DENIED DRM, their customers would NOT be shut out! DRM would simply be altered.

    The fact that the DRM is already broken means that there is nothing to gain by locking anyone out of it. HDCP is dead. The only DRM left is the input DRM, and this can all be done in software without any involvement of the GPU.

    If AMD were to lose DRM, it would be implemented in SOFTWARE, and the security of the software-DRM would be as strong as current hardware-DRM is (which is obviously not strong at all since it is broken). No customer would lose DRM. Regular consumer (i.e. non-technical) customers wouldn't even KNOW about it since these lockouts don't happen instantly. The software solution would be implemented BEFORE the DRM revocation even went through.

    And HDCP being dead? This won't make ANY difference to a consumer. As already shown, the bitrate is WAY too high for a casual consumer to deal with. In fact, this is guaranteed to have NO effect on consumer usage of DRM-infested media. The place where this MIGHT be used is in capturing/splitting/etc., the output of a satellite receiver or cable box.

  10. #120
    Join Date
    Aug 2008
    Location
    Finland
    Posts
    1,578

    Default

    Quote Originally Posted by droidhacker View Post
    Where do you get that from?
    If a MAJOR GPU vendor, like AMD suddenly was TOTALLY DENIED DRM, their customers would NOT be shut out! DRM would simply be altered.
    I wasn't talking about that. I was talking of it damaging AMD/ATi's future sales on Windows platform. Considering that platform is the main platform for desktop systems, you really have to think twice before disrupting your sales there just because of ideology issues on some minor platform.

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •