LWN.net Logo

The FSF's open letter to Google

The Free Software Foundation has laid down a challenge to Google in the form of an open letter. "With your purchase of On2, you now own both the world's largest video site (YouTube) and all the patents behind a new high performance video codec -- VP8. Just think what you can achieve by releasing the VP8 codec under an irrevocable royalty-free license and pushing it out to users on YouTube? You can end the web's dependence on patent-encumbered video formats and proprietary software (Flash)."
(Log in to post comments)

The FSF's open letter to Google

Posted Feb 22, 2010 23:52 UTC (Mon) by elanthis (guest, #6227) [Link]

Does On2/Google actually own all the patents for VP8, or are some of them
licensed?

The FSF's open letter to Google

Posted Feb 23, 2010 5:06 UTC (Tue) by eru (subscriber, #2753) [Link]

Even if Google did own the patents, and open-sourced the VP8 codec, it would then probably face the same kind of FUD assault that Theora already has trouble with: Some companies would claim there might perhaps be other patents that cover it, and therefore using it would be legally uncertain :-(

The FSF's open letter to Google

Posted Feb 23, 2010 6:36 UTC (Tue) by coriordan (guest, #7544) [Link]

That's a risk, but it might still be our best chance.

The FSF's open letter to Google

Posted Feb 23, 2010 8:34 UTC (Tue) by elanthis (guest, #6227) [Link]

You can't get around that. People can (and probably have) done the same to
every format. Threats and fact are two different things, and I'm only
interested in the fact here.

The simple question is whether (a) the FSF's request even makes any remote
sense at all or (b) a Free/Open VP8 is legally impossible due to third-party
software patent licensing.

The FSF's open letter to Google

Posted Feb 23, 2010 11:15 UTC (Tue) by bawjaws (guest, #56952) [Link]

On2's business model relied on the fact that they didn't infringe on anyone
else's patents, so yes the FSF call does make sense, and indeed freeing the
codec is about the only sensible reason anyone has come up with for Google to
have spent the 130 million dollars on this purchase.

The FSF's open letter to Google

Posted Feb 23, 2010 18:39 UTC (Tue) by drag (subscriber, #31333) [Link]

All you have to do is tell people that if Theora or VP8 infringe submarine patents then it's extremely likely that H.264 does also. The likelihood that some fundamental aspect of video compression that Theora infringes and H.264 does not infringe is pretty f-ing small.

Also Theora has been around for many year longer then H.264. H.264 has several revisions and the oldest one is only a few years old. In the time between when Theora was released and H.264 there have been literally tens of thousands (possibly hundreds) of software-released patents that have been released.

Thus if your worried about submarine patents then H.264 quite possibly is _RISKIER_.

AND....

Just because you pay licensing to MPEG-LA does NOT protect you in any way shape or form from third party patents. Companies that implement H.264 in hardware and software.. and the MPEG-LA folks themselves do NOT do patent searches to protect their customers. Patent searches only invite litigation and are extremely expensive so it's very rare that people actually do them.

Paying money to MPEG-LA only protects from lawsuites from MPEG-LA members. And only patents that MPEG-LA members have submitted to the licensing pool. Your not protected from anybody else or any other patents by paying that money. There is no chance that MPEG-LA members would assume liability on your behalf if you end up getting sued.

---------------------------------

The main source of this FUD is, of course, Apple.

Apple is in fact a member of the MPEG-LA group and does have patents that cover aspects of AAC and H.264. So the reason they are spraying Theora FUD is not because they are scared of patent infringements... it's because they make more money when you use H.264 in your product.

If H.264 was adopted as a internet standard then that would be extremely lucrative financially for companies like Apple that make money from licensing. They would not only get licensing money from all their competitors, it would be possible for them to make their own products that much cheaper since they own the patents.

Also thinking in this context... Why do you suppose Apple is fighting flash? Quite possibly one of the reasons is that if everybody uses Adobe Flash for H.264 decoding then only Adobe has to pay blanket licensing fees (which they already have done)... while if everybody has to pay H.264 licensing fees independently if they put H.264 support natively into their systems. Then that means it's likely that MPEG-LA will make much more money.

The FSF's open letter to Google

Posted Feb 24, 2010 17:15 UTC (Wed) by lgerbarg (subscriber, #57988) [Link]

If you are going to play the FUD card, you should at least fact check what you say. You are correct, Apple is a member of the MPEG-LA, and is part of the H.264 patent pool. If you want to see who holds publicly disclosed patents on H.264, you can download the list (PDF). There are over 1000 patents on the codecs by ~30 companies. Apple holds 1 patent on H.264, Microsoft hold 65, Fraunhofer hold 72, Toshiba holds 137, LG holds 198, and Panasonic holds 377 patents. Apple almost certainly does not derive any meaningful revenue from the H.264 patent pool.

I take Apple at their word on the patent issue. Think about it, there are 1000 patents by 30 companies on H.264, and there is still the risk of some submarine patent existing. Clearly Apple is exposed to that risk, but they are going to have to expose themselves to something if they want to play video, best to put themselves in the in the same boat as everyone else, and minimize the risk by not committing to anything besides what they need for their products. Even if the total risk on Theora is less than total risk on H.264, they are already committed to H.264, so Theora is just additional risk to them, and until there is market desire whose value outweighs the cost of that risk it just isn't worth it.

On2 may have claimed their codecs don't infringe on other people's patents, they may even think it, but they can't guarantee it. As an example, MS thought it owned the patents to VC-1, but when it pushed it through the standardization process for SMPTE which required the various other major video companies to disclose patents it turned out that 14 other companies had patents. The last time Apple was evaluating adding Theora support they asked Xiph if they would be willing to submit Theora to a standards body in order to force companies to disclose patents. The Xiph group actually said they would have no issue with doing that, except for the expense of the process,. That is indicative of the whole problem with Theora. People keep claiming it has no patent risk, but there are no publicly available patent searches or submission to standards bodies that would force the major companies to declare their patents, primarily because doing so is expensive. If MS ands it army of lawyers missed patents from over a dozen companies with VC-1, why should we think Theora and On2 could do better?

As much fun as it is to entirely write off a companies rationale because you don't share their concerns does not make those concerns unreasonable. Realistically, if Google wants to open up VP8 it does not improve on the patent situation relative to Theora unless they either submit it to a standards body (at which point they risk finding out there are other patent holders and end up in the same place as H.264), or offer blanket indemnification, which could be unbelievably expensive for them. I suspect it is pretty likely they will submit it to the standards bodies, but if they do it may take a year or two before it is clear whether or not they even own the all the patents.

The FSF's open letter to Google

Posted Feb 24, 2010 17:25 UTC (Wed) by foom (subscriber, #14868) [Link]

This is the most cogent explanation of the situation I've read thus far. Thanks!

The FSF's open letter to Google

Posted Feb 23, 2010 9:35 UTC (Tue) by bangert (subscriber, #28342) [Link]

But Google is in a position where it can generate facts: by using VP8 on youtube. This would instantly make VP8 the defacto standard of online video codecs - and that will stop the FUD machine right there. Either by way of proving that the lawsuits dont happen or because it does indeed attract such a lawsuit.

This is why the FSF also urges Google to start using it on youtube and why its important that Google opens up the format. If it isn't open, its just the same thing as H.264. Actually its worse than that, since its controlled by a single entity an cant be used royalty free until 2016.

Given that it will be dificult to create revenue off VP8 when it remains closed, it makes a lot of sense for Google to spend that kind of money to open up the online video market.

This, once again, will be a complete game changer - this time with Adobe (and to a lesser degree windows media, real and quicktime - but they dont matter much anymore anyway) being on the loosing end. Google has done it in the past and this wont be the last time...

What's next?

The FSF's open letter to Google

Posted Feb 23, 2010 18:40 UTC (Tue) by Max.Hyre (subscriber, #1054) [Link]

[I]t would then probably face the same kind of FUD assault that Theora already has trouble with: Some companies would claim there might perhaps be other patents that cover it[.]
Anyone who falls for that doesn't understand the patent problem. The same argument applies to every other codec out there. Nay, to every single piece of software out there. Just because you're a corporation selling proprietary software doesn't mean you have detected (or even can detect) every patent it violates. Ask Microsoft.

The FSF's open letter to Google

Posted Feb 23, 2010 23:30 UTC (Tue) by linusw (subscriber, #40300) [Link]

So they FUD:ed about Ogg Vorbis as well and now everybody and his dog is using Vorbis without being sued. (Several asian MP3 players, Spotify, every Android device.)

The FSF's open letter to Google

Posted Mar 1, 2010 16:51 UTC (Mon) by Hanno (guest, #41730) [Link]

Samsung sells products with OGG support. And they haven't been sued so far, as far as we know.

If Samsung isn't already big enough to activate a patent troll's submarine patent, I wonder who is.

A bit off topic...

Posted Feb 23, 2010 0:49 UTC (Tue) by yokem_55 (subscriber, #10498) [Link]

But can anyone explain the FSF's rationale for supporting the development of Gnash? While it's nice to have a free software runtime for flash, 98% of the swf's that gnash will ever execute are proprietary applications themselves. So in the end we'll have a Free Software runtime that will only ever be used to run non-free apps and will make those non-free apps available to more platforms than before. I suppose it makes for a better situation than just running Adobe's runtime, but it still seems a bit odd that the FSF is backing the development so much.

A bit off topic...

Posted Feb 23, 2010 1:04 UTC (Tue) by JoeBuck (subscriber, #2330) [Link]

If Flash is basically a required part of the web, then a free platform has to have support for it. If, on the other hand, the industry moves away from Flash, Gnash becomes less important.

And for all practical purposes Flash is required; my nontechnical wife and daughter are OK with running Linux because Flash works.

I said "Flash works", not "Gnash works". Gnash will always be playing catch-up.

A bit off topic...

Posted Feb 23, 2010 2:14 UTC (Tue) by Burgundavia (guest, #25172) [Link]

By that logic, we should develop absolutely no free software, merely because it might be used to run/interpret close source applications.

A bit off topic...

Posted Feb 23, 2010 7:05 UTC (Tue) by johill (subscriber, #25196) [Link]

But compare to their stance on Mono.

A bit off topic...

Posted Feb 23, 2010 8:37 UTC (Tue) by Darkmere (subscriber, #53695) [Link]

That Mono is a good thing to run existing applications and have compability, but shouldn't be encouraged for new development because of other situations re: patents and licensing and control?

A bit off topic...

Posted Feb 23, 2010 17:17 UTC (Tue) by trasz (guest, #45786) [Link]

May be worth mentioning that the patent and licensing problems are just a FUD, not backed by anyone not affiliated with FSF.

A bit off topic...

Posted Feb 23, 2010 23:42 UTC (Tue) by nix (subscriber, #2304) [Link]

So Novell entered a patent covenant with Microsoft because the FSF asked
them to? That's the oddest interpretation of those contentious events that
I have *ever* heard.

A bit off topic...

Posted Feb 24, 2010 6:45 UTC (Wed) by trasz (guest, #45786) [Link]

No, but their deal with Microsoft was not really related to Mono.

A bit off topic...

Posted Feb 23, 2010 11:40 UTC (Tue) by Los__D (guest, #15263) [Link]

They think that using C# is a great risk, but does not think that using C# is unethical.

They even have their own implementation, DotGNU Portable.NET.

I really can't see any difference in stance.

http://www.fsf.org/news/dont-depend-on-mono
http://www.gnu.org/software/dotgnu/

A bit off topic...

Posted Feb 23, 2010 15:42 UTC (Tue) by johill (subscriber, #25196) [Link]

Yeah you're all right of course, I got confused there for a moment and posted while so :)

A bit off topic...

Posted Feb 23, 2010 17:16 UTC (Tue) by trasz (guest, #45786) [Link]

GNASH allows people running "Free" operating systems to use Flash - which is rather crucial part of the Web - without depending on closed-source software, which is good for said systems and doesn't really matter for closed source.

On the other hand, Mono could allow people to run the same desktop or server applications on whatever system they choose - either "Free" or not. And, from the "Freedom" point of view, this is bad - because then people can easily use "Free" applications under other operating systems, removing one of the possible reasons to switch to the "Free" operating system. See RMS opinion on porting KDE to Windows.

For Java it doesn't matter, because Java doesn't matter on the desktop.

A bit off topic...

Posted Feb 23, 2010 4:03 UTC (Tue) by eklitzke (subscriber, #36426) [Link]

I'm obviously not a lawyer, but isn't this true for Javascript (or any other web-based scripting language) as well?
  • Even though Javascript is distributed in source form, the JS programs you download and execute are essentially proprietary unless licensed otherwise. It's probably not legal to copy someone else's JS program if they haven't explicitly authorized you to do so.
  • Minified JS can be extremely inscrutable. The Closure compiler (what Gmail, Google Docs, etc. use) is a great example of this — it will do whole program analysis and rename functions and symbols, so not only are the variables and control structures obfuscated, the exported function names are as well. I know the FSF or RMS have written about this before.
Now clearly Actionscript bytecode is more opaque than JS (even when it's minified), but the fact remains that most JS isn't licensed freely or necessarily hackable.

A bit off topic...

Posted Feb 23, 2010 6:39 UTC (Tue) by coriordan (guest, #7544) [Link]

RMS also wrote about this javascript problem last year: http://www.gnu.org/philosophy/javascript-trap.html

A bit off topic...

Posted Feb 23, 2010 4:41 UTC (Tue) by BrucePerens (guest, #2510) [Link]

And any Free web-browser is intended to display proprietary web pages most of the time. They aren't under GPL, or Creative Commons, they aren't in the public domain, and they generally say nasty things like "all rights reserved" somewhere at the bottom of the page.

A bit off topic...

Posted Feb 23, 2010 6:03 UTC (Tue) by njs (guest, #40338) [Link]

In fact, it's so common that "All rights reserved" is the most common sentence in the English language. We have a ways yet to go...

The FSF's open letter to Google

Posted Feb 23, 2010 2:38 UTC (Tue) by dkk (subscriber, #50184) [Link]

I find it hard to believe On2 was bought by Goog just for them to sit on the technology, which in my opinion makes the FSF's open letter kind of irrelevant.

The FSF's open letter to Google

Posted Feb 23, 2010 3:03 UTC (Tue) by clump (subscriber, #27801) [Link]

The letter and even the blurb that links to said letter do aid greatly in understanding the FSF's rationale. Shocking, that.

The FSF's open letter to Google

Posted Feb 23, 2010 12:05 UTC (Tue) by csamuel (✭ supporter ✭, #2624) [Link]

Sylvia Pfeiffer has a good blog post on some of the issues around open
sourcing the VP8 codecs that Google may face should they decide to take
this route, not least that it is purely a video codec so it is no silver
bullet.

She also makes the point that Google could assist Theora by renegotiating
the license with Xiph for VP3 to clarify the status of independent
reimplementations of it.

http://blog.gingertech.net/2010/02/20/googles-challenges-...

The FSF's open letter to Google

Posted Feb 23, 2010 18:42 UTC (Tue) by JoeBuck (subscriber, #2330) [Link]

I think that she risks asking too much. Google can't be expected to indemnify everyone in the world against any submarine patent that anyone else might hold, even if they go after an independent implementation and not after the code that Google releases. That's asking them to open their corporate treasury to every ambitious patent lawyer in the world. Since Google isn't going to do that, the FUD merchants can then point to that blog post and say it's just too risky to use something other than the "safe" H.264.

There's a mistaken belief that alternative codecs are uniquely vulnerable to submarine patents, while if you go with H.264 you're safe, because you've licensed the patents. You aren't safe; some other troll might have a patent that's not in the pool, and the licensing body does not indemnify you.

The FSF's open letter to Google

Posted Feb 24, 2010 14:55 UTC (Wed) by forthy (guest, #1525) [Link]

If being able to pay for the patent is resolving the FUD problem, then Google should just make companies pay for the VP8 patents - if they want to. A lot of company bean-counters believe in paying for things, and then owning them, and they have great troubles with gifts, so simply sell them what they can have for free. Could be a complete bundle of patents plus GPL exemption (if VP8 is released under GPL) plus maintenance.

The FSF's open letter to Google

Posted Feb 24, 2010 15:31 UTC (Wed) by bawjaws (guest, #56952) [Link]

That's not what she's asking for.

On2 had some patents on VP3, which forms the basis of Theora. They fully
intended to make these patents available for anyone to use, and that's
apparently what the legal documents say. Furthermore, Monty from Xiph says
the patents aren't even necessary for Theora, they only apply to specific
implementations.

Apple, as part of their reasons for not using Theora, claimed that the
patent-grant was too narrow and only covered implementations derived from the
VP3 source code, and not independent implementations.

This appears to be a misreading of the situation but either way Google could
provide some clarity by saying "These patents we bought were already made
available for anyone to use by On2 but just to be clear and avoid any doubt
we assert that they are still available for use" in whatever level of
legalese is required for this kind of thing.

This may be pointless, because Apple may be just blowing smoke and trying to
find fault and they'll fall back on other arguments, but it shouldn't cost
Google anything but some lawyer time. It's more about appearances than
facts and won't expose Google to any kind of patent risk.

The FSF's open letter to Google

Posted Feb 24, 2010 16:07 UTC (Wed) by bawjaws (guest, #56952) [Link]

Apologies, she does actually mention indemnification, but is (as clarified in
the comments) raising it as what people who have other reasons for opposing the
codec would say, not something she is actually requesting of Google.

She suggests that this is not wise or likely to happen but that standardisation
would be route to achieve something similar to this, at least to whatever
degree other standardised codecs reach, necessarily limited as that might be.

Copyright © 2010, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds