Slashdot Mirror
Google Chrome's Inclusion of FFMpeg Vs. the LGPL
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."
It's not surprising that Miguel wants in on this.
What projects has he worked on in the past decade that didn't revolve around patents by Microsoft or others?
Mod me down, my New Earth Global Warmingist friends!
So, what you're telling us is that although all of us in the States have been using FFMpeg illegally without a license for years, when someone finally decides to try and be legal and purchase a license, now they are still in the wrong?
Sounds like the FFMpeg people need to start dual-licensing or something - from what I can tell they are OK with people obtaining licenses to use FFMpeg.
Of course, I'm still missing the biggest point here - since when do they need FFMpeg for HTML 5 support? It doesn't require any patented codecs, and they could always use DirectShow filters.
It isn't when it's inside of Chrome.
The LGPL expressly allows closed-source and even non-free-as-in-beer software to link to an LGPL library, either statically or dynamically, without violation of its terms. That's what makes it lesser than the GPL.
I agree with sentiment in the last link though, this is none of the FSF's business -- the FFMPEG people are the only ones that can claim to be aggrieved here. Until that happens, this is much ado about nothing since there can be no violation of license terms to which the holder of the copyright does not object.
I should probably clarify that Google didn't say *which* libraries would be used, but I assume they were referring to Windows Media Player for Windows, Quicktime for Mac, and GStreamer for Linux. Someone in the know can probably confirm or deny.
Javascript + Nintendo DSi = DSiCade
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.
How we know is more important than what we know.
I must say, it seemed more than a little ... odd ... for the founder of a completely and utterly proprietary competitor to post off-topic messages to a mailing list trying to probe his direct competitor on their adherence to a free software license.
http://rocknerd.co.uk
Is there really an issue here? Is it impossible to freely redistribute Chrome? According to this page:
http://code.google.com/chromium/terms.html
It seems that anybody can redistribute the code and/or binaries, with the possible exception of the parts that are trademarked (similar to Mozilla).
I noticed that google does give credit to ffmpeg on its "about" page, funnily enough I checked it out last night and clicked the ffmpeg link to read about what they had to say! So I think it is now included.
like phosphorescent desert buttons singing one familiar song
I have to reserve judgement until I know the details of Google's patent deal. It is possible that it includes downstream use.
All patent licenses are not the same. Some are a per-unit royalty, others are a lump-sum purchase of the use of the invention. It's even possible that Google did negotiate "downstream redistribution" of the library if they paid enough cash.
From the attached article I can gather that since FFMpeg uses the LGPL 2.1 (not 3.0) that their obtaining a third party license for something else that prohibits them from granting similar rights for that bit of code does not affect their ability to grant rights for use the FFMpeg libraries. As they put it:
The fact that Party B may have a patent license with an unrelated third-party is irrelevant as long as it doesn't prevent Party B from granting people the rights LGPL 2.1 requires they grant them (namely, only those rights it in fact received from Party A).
Again this all seems rather moot anyway. A lot of operating systems these days include FFMpeg libraries as well as the H.264 and AAC libraries (which is really what this is all about). I know people feel like the idea of linking native libraries from the OS (which may or may not be there) goes against the universality of the HTML5 video/audio spec (and I can't say I disagree), but it would seem that for something as ubiquitous and freely licensable as the FFMpeg libraries, this argument is a bit overblown.
Science will save us. The question is, will it destroy us first?
Nope, you've got it wrong. Chrome includes ffmpeg.
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2009-June/020035.html
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. I think that, right or wrong, Google's probably on solid legal ground there. They didn't introduce the patented code into the library, they didn't create the library, so I don't think the law'll have much trouble allowing them to redistribute the code under the same terms and with the same rights as they got from it's creator.
I think it's a situation the GPL and LGPL don't contemplate explicitly. The situation where Google was adding the infringing code to a library they received under (L)GPL terms and then redistributing the results, that's exactly what the v2 language covers. But I'm not sure even the v3 language covers the situation where the holder of the patent license isn't the one who put the infringing code into the library and where the person who did put it in doesn't hold a patent license and has no rights to grant downstream recipients. If the library was under LGPL v3 I think you could make an argument that the automatic grant of rights in paragraph 11 kicks in, since Google does hold a relevant patent license, and that if their license doesn't let them do what paragraph 11 requires then they can't redistribute, but LGPL 2.1 doesn't contain anything explicit corresponding to v3's paragraph 11.
I think Google's right here, it does in fact come down to what the patent-holder says. If they sue Google and get an order blocking Google from distributing infringing code, then Google can't distribute the library. If the patent holders sue the library's author and get an order blocking distribution of the library without a patent license, Google can't distribute the library. But until then, Google can't be forced to add any rights that they didn't get when they received the library.
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.
Warning: this article may contain humor, sarcasm, parody, and perhaps even irony. Read at your own risk.
IANAL (but I try to keep up on patent/copyright law), but here's how I think that would go:
gr8_phk: Here you are, sir. One compiled binary of FFMPEG, with source!
customer: Thanks!
patent_lawyer: Hold on there! You don't have a patent license; pay up gr8_phk!
gr8_phk: I don't need a license, Google gave me one since I got this off of Google.
patent_lawyer: Google didn't give you shit. Pay up!
Google: He's right, we didn't give you anything.
gr8_phk: Grr!
FFMPEG developers: Wait a tick there Google! You can only use our code if you give everyone who got the source from you a patent license. ...
Google: Well, that isn't the agreement we have with the patent holders. Sorry.
FFMPEG developers: Fair enough, our lawyers will be suing you for copyright infringement.
Google: Ha! You're going to sue us? I doubt it. We'll tie this up in court for years until you throw in the towel.
FFMPEG developers:
Stop making sense, AKAImBatman. You're gonna spoil all the fun of this thread before it even gets going.
We need to destroy Google immediately! After all, what have they ever done for us?
You are welcome on my lawn.
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.
Wow, a capital letter...
http://bash.org/?367896
I think I understood it better before I read that.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
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.
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.
All your base are belong to Wii.
> Oddly, I have Chrome 2.0.172.30, but no FFMPEG license in sight.
That's because <video> support was added in Chrome 3.x
Slashdotters think copyright law is wrong and piracy is okay.
I think you have two distinct groups of slashdotters confused. Those of us who support the GPL do not pirate Microsoft Windows (we don't want it or need it) and we do not pirate music.
We're fundies man. Do not make me car bomb your OEM Microsoft Windows CDs.
How do you know isn't just a collective of Lisp Daemons at this point?
You're misinterpreting the LGPL. The language quoted by /. above says
Now, imagine it this way:
Let A be some 3d party software covered by patents that is unrelated to FFMpeg. Let B be Chrome. Let C be FFMpeg. Google got a patent for A. The patent is unrelated to FFMpeg.
The LGPL says that if you cannot distribute C without patent royalties, then you cannot distribute C at all.
People are making one of two mistakes regarding this issue. They're either assuming that the patent Google has cover C (this is thanks to Slashdot's shitty summary that makes it sound like this, but Google points out quite clearly this is not the case) or they think the LGPL says that if any patent restricts the distribution of A contained in B, then you distribute C if it is contained in B. This is a poor interpretation of the above-quoted LGPL clause.
The last link I provided (http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-June/020035.html) explains better than I the stance Google has taken. I think this stance is correct.
Disclaimer: I am merely a law school graduate. I have not taken the bar yet. I am not a lawyer. I am not your lawyer. This is not legal advice.
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.
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.
Regardless of the legality of what Google is doing, the point that only the ffmpeg folks are able to enforce the LGPL is still significant. Google's relationship with the project is likely quite good. They're sponsoring 9 students as part of their Summer of Code program. If one of the contributors were to file suit, it's likely that other project members could persuade that person to drop the suit.
Even if they are in violation of the letter of the law, they're not really in violation of the spirit of the law. They're giving back to the open source community by releasing the source to their browser. And they're paying to add new functionality to ffmpeg. The only issue is that their legal team felt the need to cover the company by purchasing a license. And they would have been foolish not to, since the threat of a LGPL lawsuit is much less than a patent infringement lawsuit. AFAIK, even if they were to lose, they would be given a chance to come into compliance. And this only becomes an issue if someone who contributed to ffmpeg feels that this minor issue merits the hassle of a lawsuit and probably end any GSoC sponsorship for the project in the future...seems unlikely to me.
So the point that only the ffmpeg contributor have standing to attempt to enforce the LGPL seems pretty important since it likely means that no one with the right to do so will go through a ton of hassle to iron out a few legal details when the company has been nothing but gracious towards the project as a whole and even towards the open source community as a whole. There's far too many companies violating the (L)GPL that are acting in bad faith attempting to leach off the open source community that would make better targets. ffmpeg even maintains a list of such companies on their site.
"Don't blame me, I voted for Kodos!"
Mpeg AFAIK is a freely implementable spec
This is not true. MPEG is covered by a variety of patents which may or may not be valid in a number of different countries.
"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.
This is not a complicated issue at all. Either ffmpeg requires patent license to use/redistribute or it doesn't. If it doesn't - no problem, nobody is in violation.
If it does then Google doesn't have a license to redistribute but neither do the authors! Hence, Google are supplying the source code under exactly the same terms so there is no ground for complaint for ffmepg's authors. The fact that Google, lets say for their own peace of mind, bought a license to use any relevant patents is frankly irrelevant and none of anyone else's business.
Exactly.
It should be realized that the party who would enforce any such breach of copyright would be the people who hold copyright: its writers, whereby any suit on a breach of that clause would have to argue that there exist valid, applicable patents that apply to the capabilities GPL licenced code,. a stance copyright holders have not taken. [1]
http://www.ffmpeg.org/legal.html
He's not being misleading, he's just not telling you everything. They are distributing the library without an attached patent license. *They* have a patent license for the library. The question is whether or not distribution requires the patent license.
The heart of the issue is whether or not the functionality is used in the redistribution. I'm quite certain his take is that it isn't. The *running* of the application is what uses the functionality. Thus the user requires a license, not the distributor. Google is redistributing the library, but clearly they intend it to be "as is" without a patent license. If the user feels they need a patent license then they need to get one.
This is a weakness of V2.1 of the LGPL. Contrast the wording in V2.1 with that in V3 (section 11) and you can see what's causing the problem. In V3 you have the clarification, ""Knowingly relying" means you have actual knowledge that, but for the patent license, your conveying the covered work in a country, or your recipient's use of the covered work in a country, would infringe one or more identifiable patents in that country that you have reason to believe are valid." In other words it relies on the patent if the *user's* use of the software requires the patent.
Basically Google are covering their ass and simultaneously going against the spirit of the LGPL. But they are probably in the clear legally. Personally I think they are being weasels.
1) Give software away without restriction. Result: someone takes the code, incorporates it into their paid for product and using the revenue from their near monopoly blows the 'free' version out of the water. No more free version.
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 try and argue semantics simply because this won't be in support of your premise, it's true that they no longer have 100%; but they still wouldn't be that far below 90%.
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. Thus, it could work for them, whether or not it couldn't work for a smaller startup.
Even with small startups, the whole point isn't to necessarily use the BSD license for everything. Use it for things that ought to have unencumbered reference implementations; video codecs are a good example. That way, it getting copied by other companies is entirely the point; some of said companies will add value to it (and close it) if they want, while others can also develop it but keep it open as they choose, while still others can rely on availability of at least the reference baseline.
Corporations are entirely free to make closed, modified derivatives of Apache; again, if Stallman is right, why haven't they been destroyed yet?
Your entire argument is fear-based; as all of the FSF's rationale is. If the, "corporations are going to destroy everything!" argument is valid, how come any of the BSDs still exist at all? Come to that, why didn't Microsoft destroy the WC3 after they acquired Internet Explorer? How come virtually any of the Internet's protocols still exist, rather than single, monopolised implementations? As I said above, Apache isn't licensed with the GPL; how come it hasn't been destroyed yet?
I took "intellectual property lawyer" to mean works in the complex area of law surrounding contracts and licencing, and the interpretation and supposed violation of the above.
Yay me!
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.
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.
Bye!