Forgot your password?
typodupeerror
Patents Businesses Google Graphics Software The Internet News

Google Chrome's Inclusion of FFMpeg Vs. the LGPL 245

Posted by timothy
from the to-make-one's-head-spin dept.
An anonymous reader writes "Google has recently added FFMpeg to Chrome to better support HTML5's video element. FFMpeg is licensed under LGPL 2.1, which states that 'if a patent license would not permit royalty-free redistribution of the Library by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Library.' Google admits to having obtained a patent license for their use, but still claims they are not violating LGPL. Among the confused we find Håkon Wium Lie and Miguel de Icaza, who wonders what FSF might say. Google doesn't feel like asking FSF for clarification."
This discussion has been archived. No new comments can be posted.

Google Chrome's Inclusion of FFMpeg Vs. the LGPL

Comments Filter:
  • And it doesn't (Score:5, Insightful)

    by QuantumG (50515) * <qg@biodome.org> on Sunday June 07, 2009 @08:17PM (#28245551) Homepage Journal

    if a patent license would not permit royalty-free redistribution of the Library by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Library

    See that word "if"? The patent license permits royalty-free redistribution of the Library... so it's not an issue.

    Similarly, we've heard nothing from the authors of the Library - you know, the copyright owners, the only ones who have any legal standing? So maybe the peanut gallery should shut the hell up already.

  • by Anonymous Coward on Sunday June 07, 2009 @08:22PM (#28245583)

    Producing legal spagetti crapazola like that should be punishable by hanging by the nuts. F*** GPL. Go BSD, go free.

  • Re:And it doesn't (Score:5, Insightful)

    by mrsteveman1 (1010381) on Sunday June 07, 2009 @08:24PM (#28245599)

    Internet rule #28854, everyone has a right to complain about everything.

  • Re:And it doesn't (Score:2, Insightful)

    by Stumbles (602007) on Sunday June 07, 2009 @08:41PM (#28245753)
    If the peanut gallery did shut the hell up.... Slashdot would not exist. Come to think about it, Slashdot is the pachyderms heaven.
  • Re:And it doesn't (Score:5, Insightful)

    by QuantumG (50515) * <qg@biodome.org> on Sunday June 07, 2009 @08:57PM (#28245837) Homepage Journal

    For branded encoder and decoder products sold both to end users and on an OEM basis

    Reading comprehension, you failed it.

  • by John Hasler (414242) on Sunday June 07, 2009 @09:01PM (#28245861) Homepage

    They are distributing the library under the terms of the LGPL with no additional restrictions and so are complying fully with the license. Whether or not they are violating their patent license by doing so is their problem.

    The situation this clause of the LGPL is aimed at is one wherein Google would be obligated by their patent license to require that everyone they distributed the program to sign a patent sublicensing agreement that took away rights granted by the LGPL.

  • by Mihg (2381) on Sunday June 07, 2009 @10:03PM (#28246215)
    On the other hand, DannyB is an intellectual property lawyer, and you aren't. Furthermore, "the ffmpeg folks" would include "any contributor to ffmpeg", so your point is moot.
  • by Jiro (131519) on Sunday June 07, 2009 @10:22PM (#28246337)
    Anyone who has copyright in a work cannot violate their own license. They could create a license which says "you can only distribute this if you compute pi to the last digit". Anyone who receives it wouldn't be allowed to distribute it (since the requirement is impossible), but they, being the creator could distribute it just fine. The license only restricts other people.

    If ffmpeg is under a license which says, basically, "you can only distribute this if you can pass on an impossible patent license", the creator can still distribute it without a patent license. They would, of course still be violating the patent, but they wouldn't be violating the copyright. A third party *would* be violating the copyright as well as the patent, and could be sued for copyright violation (permission to distribute only under impossible conditions means no permission).

    <i>So either it's creator could grant them an LGPL 2.1-compatible patent license, or the library can be distributed without a license, or it's creator couldn't have legally distributed the library to them.</i>

    The creator can always legally distribute the library to them (with respect to copyrights), even if the copyright license is impossible.
  • by Score Whore (32328) on Sunday June 07, 2009 @10:47PM (#28246503)

    I think Google's relying on a technicality, but it's a significant one. In this case Google isn't the creator of the library, they received it from it's creator. So either it's creator could grant them an LGPL 2.1-compatible patent license, or the library can be distributed without a license, or it's creator couldn't have legally distributed the library to them.

    You're confused. The author of the product doesn't have to abide by the license, they own the copyright and can do anything they want. The LGPL doesn't apply to them. It's perfectly legit for them to say, "hey, here is this code that implements patented algorithm X. if you want to use it you'll have to get your own license from the patent holder."

    As far as this goes with the ffmpeg authors violating the patent by implementing this stuff in the first place, there is a certain amount of protection from patent litigation if you are doing research. Not as safe as having a license, but better than selling a blue-ray player without one. Additionally these developers are probably pretty close to judgment proof (ie. they mostly have no money to pay any judgments against them.)

  • by msuarezalvarez (667058) on Sunday June 07, 2009 @10:48PM (#28246509)
    It is unhelpful to a lot of people, probably. But writing software to follow the particularly idiotic US law in this regard ends up being unhelpful for way more people. Reflect on the fact that the whole US population is, in a global perspective, only a very loud minority...
  • by Hal_Porter (817932) on Sunday June 07, 2009 @11:01PM (#28246575)

    The Chrome people say that you're getting a patent license for H.264, etc. if you use Chrome. Fine.

    The interesting question is "Does my patent license for H.264, etc. extend to any decoding, or only that done by Chrome?". Said in another way, is my patent license only good if I'm doing the decoding in Chrome or does it apply to decoding done by me? If it is the latter, then anyone who wants a patent license can just download Chrome -- now they have a free patent license.

    I think the fact that if the patent license applied to decoding done by software other than Chrome then everyone could get a patent license strongly implies it doesn't. I.e. the patent holders won't sue Google and they won't sue users of Google's software. If those users use other software that decodes H.264 I can't believe that the patent holders have agreed not to sue them.

  • by QCompson (675963) on Sunday June 07, 2009 @11:06PM (#28246607)
    Christ dude, if you're going to go car analogy, you have to go full-car. Leave out FFmpeg and MPEG and all that other confusing crap and just talk about cars.
  • by i.of.the.storm (907783) on Sunday June 07, 2009 @11:18PM (#28246689) Homepage
    Not sure why this was modded troll, since it's factually correct. Chromium is open source, yes, but Chrome itself is not. It's a fine distinction, and I'm not sure what parts of Chrome differ from Chromium, but I think the automatic updater service is not installed with Chromium, among a few other things. I think Chromium also lacks Google branding, probably for sticky copyright issues. Of course, the parts that use FFMpeg are probably open source, although we can't say for certain since the Chrome source is not available, but there's no reason why they should be different in Chrome than Chromium.
  • by Anonymous Coward on Sunday June 07, 2009 @11:24PM (#28246731)

    The Chrome people say that you're getting a patent license for H.264, etc. if you use Chrome. Fine.

    Doubtful. Google is getting a license to distribute patented technology to you.

  • by MaskedSlacker (911878) on Monday June 08, 2009 @12:10AM (#28246997)

    What's to stop me from writing a program that makes use of the copy of the codec installed by Chrome? I'm only using what the patent license said I could...

    While you might be right, it would be unenforceable, and therefore irrelevant and meaningless. Kinda like the injunctions against DeCSS.

  • by TheoMurpse (729043) on Monday June 08, 2009 @01:28AM (#28247361) Homepage

    I didn't know FFMpeg could afford to employ a team of lawyers for years. Note that they cannot use FSF lawyers, because the FSF requires that you assign all copyright over to them before they will defend you.

    Note: I think that is true. I recall reading something like that on the FSF site years ago, but take it with a grain of salt.

  • by TheoMurpse (729043) on Monday June 08, 2009 @01:40AM (#28247415) Homepage

    Anyone who has copyright in a work cannot violate their own license.

    That's not true. It's possible to lock yourself out of the right to use your own copyright via a license through really shitty drafting of the license (or on purpose). If I recall correctly from my software licenses class (and I might not), the magic word "exclusive" is all that is needed in the grant clause and BAM, you can't use your own copyrighted work!

    I am merely a law school graduate. I haven't taken the bar yet. I'm not a lawyer. I'm not your lawyer. This is not legal advice.

  • by jipn4 (1367823) on Monday June 08, 2009 @01:53AM (#28247461)

    It matters what patents exist. If it is the position of the FFMpeg authors that the patent license that Google has obtained is actually required for royalty-free distribution, then nobody can redistribute FFMpeg at all.

  • by WillKemp (1338605) on Monday June 08, 2009 @02:38AM (#28247679) Homepage

    This license is for the sole purpose of enabling you to use and enjoy the benefit of the Services as provided by Google, in the manner permitted by the Terms.

    Yeah, so...? That's just about the purpose of the licence, it doesn't say anything about how you can or can't use the software that's being licensed. If it said something like "you may use the software for the sole purpose of enjoying the benefit of the services...." it might be relevant. The whole paragraph is badly drafted anyway. "...license to use the software provided to you by Google as part of the Services as provided to you by Google..." is ambiguous.

  • by Chrisq (894406) on Monday June 08, 2009 @04:11AM (#28248217)

    Reflect on the fact that the whole US population is, in a global perspective, only a very loud minority with an extremely large economic footprint.

    When you look at it from that angle, the population figures become much less relevant to the discussion.

    For a commercial organisation maybe, but for open-source programmers in Europe no way. They will not say "I won't implement this interesting and useful codec because people in the USA might have legal problems using it" any more than Zimmermann would have said "I won't produce PGP because it may cause legal problems for people in France using it".

  • by GauteL (29207) on Monday June 08, 2009 @04:31AM (#28248337)

    "I didn't know FFMpeg could afford to employ a team of lawyers for years. Note that they cannot use FSF lawyers, because the FSF requires that you assign all copyright over to them before they will defend you."

    I'm willing to bet that there would be at least some organisations willing to help fund this lawsuit. The EFF is a good bet.

  • by TheRaven64 (641858) on Monday June 08, 2009 @07:31AM (#28249243) Journal
    No, it's very helpful. Those of us outside the USA get to use the software, those of you inside the US get to go to your respective representatives and say 'look at these people outside the USA who can use this software and gain a competitive advantage over me! Change the law so that I can use it or we will not be able to compete and your tax revenues will drop!'
  • by DragonWriter (970822) on Monday June 08, 2009 @11:55AM (#28251809)

    They are violating the spirit of the GPLv3,

    Since the issue is their use of a library licensed under the LGPL, version 2.1, even if violating the "spirit" of a license was a problem, the "spirit" of a completely different license than the one applicable to the software they are using wouldn't be.

  • by ADRA (37398) on Monday June 08, 2009 @01:08PM (#28252837)

    I don't necessarily disagree with what you're trying to convey, but your approach needs some work:

    "Microsoft would tend to disagree with you, I'm sure. They managed to corner the market with a free product, in the case of Internet
    Explorer. Before you also start bleating at me about how they were leveraging their monopoly, realise how well Google are doing in the market on search."

    Netscape was an ok web browser and when IE made it to around 4 or 5, they became ok as well. The difference was that IE -was- installed by default on the majority of new desktops (mshtml cannot be uninstalled) and Netscape wasn't good enough to justify users going out to find a new one. This was during a period where more and more casual PC users were entering the PC market to use the Internet. Many of those users would care less what 'browser' they used.

    "some of said companies will add value to it (and close it)"
    Lets say that Microsoft took vorbis, added some proprietary and hence incompatible hooks to the product then made it part of the default Windows installation, nobody would cry bloody murder? I think the most likely case would be that the 'market' would split with some using microsoft's default broken implementation while others would go with the free open version. Same things happened with web development and Java. In both cases Microsoft embraced and extended to the point of breaking inter-compatibility.

    "Corporations are entirely free to make closed, modified derivatives of Apache; again, if Stallman is right, why haven't they been destroyed yet?"
    Apache is still alive because of the people contributing to it. If a closed source product (like MS) wanted to embrace and extend Apache, that's their decision. But apache is itself an implementation of a specification, not the specification itself. The only benefit someone has of taking Apache code would be to save development costs in rolling out their own Web server. I'm sure there are a ton of companies that have done just that. If your point is to develop code that people will use, mission accomplished. If your goal was to foster your personal philosophy of software without commercial value, you're probably on less steady grounds.

    "how come any of the BSDs still exist at all?"
    BSD is a bad example, since its lost favor for the vast majority of developers. Linux could be considered the slightly younger, more attractive platform to work on, so people jumped ship to Linux, but nothing stops BSD from living on forever. That's not to say its a success story in modern times vs. Apache which still maintains a very high relevance.

    "why didn't Microsoft destroy the WC3 after they acquired Internet Explorer?"
    They had a monopoly but not exclusive monopoly on the market. That's why they embraced and extended their browser to include proprietary hooks like ActiveX and many non-standard coding extensions. If given enough time, they would've completely tossed out WC3 from consideration. Lucky for us, Mozilla took over the sinking Netscape product line and saved our ability to choose web platforms.

    "Internet's protocols still exist, rather than single, monopolised implementations?"
    IPv4, TCP, UDP, all De-facto standards that EVERYONE uses. Maybe you should pick a better example, because this one is 100% inaccurate. If you're trying to come up with 'proprietary' protocols, you would be best to look at Instant Messangers, some streaming media, some email provisions and SMB. Beyond that, I can't recall many protocols that don't have proprietary hooks that limit their use. The major reason why protocols aren't proprietary is because the protocols aren't copyrightable in many(all?) countries. If its not illegal to copy, someone could analyse and reproduce said function and publish specifications for how it works without legal reprocussions. Once that's done, the protocol can be implemented by any who care to use it.

Never say you know a man until you have divided an inheritance with him.

Working...