Advertisement
Front, Kernel, Security, Distributions, Development. See your byline here on LWN.net.
Weekly Edition
Return to the Announcements page
| |||||||||||||
|
| |||||||||||||
Hold The Celebrations; H.264 Is Not The Sort Of Free That Matters (ComputerWorld UK)
Over at ComputerWorld UK, Simon Phipps says there is nothing to celebrate in the recent announcement [PDF] that MPEG-LA will not charge royalties on "web uses" of the H.264 codec for the remaining life of the patents it administers. "First, the H.264-format video needs to be created - but that isn't free under this move. Then it needs to be served up for streaming - but that isn't free under this move. There then needs to be support for decoding it in your browser - but adding that isn't free under this move. Finally it needs to be displayed on your screen. [...] The only part of this sequence being left untaxed is the final one. Importantly, they are not offering to leave the addition of support for H.264 decoding in your browser untaxed. In particular, this means the Mozilla Foundation would have to pay to include the technology in Firefox." He also posits that MPEG-LA may try to join forces with Oracle and Paul Allen's Interval Research to create a three-way patent attack on Google—this time against WebM.
(Log in to post comments)
Hold The Celebrations; H.264 Is Not The Sort Of Free That Matters (ComputerWorld UK) Posted Aug 31, 2010 15:28 UTC (Tue) by stumbles (guest, #8796) [Link] ..."MPEG-LA may try to join forces with Oracle and Paul Allen's Interval Research to create a three-way patent attack on Googlethis time against WebM. " Just a good example why software patents do not promote technologies.
Hold The Celebrations; H.264 Is Not The Sort Of Free That Matters (ComputerWorld UK) Posted Aug 31, 2010 16:47 UTC (Tue) by TeDiouS (guest, #67602) [Link]
Hear, hear!
FUD Posted Aug 31, 2010 17:11 UTC (Tue) by fandingo (subscriber, #67019) [Link]
It's interesting that Simon Phipps, and most people that have similar arguments, talk about the FUD that MPEG-LA is trying to cast on WebM. Everyone seems to hate MPEG-LA, and thinks that patents will be the end of (video, html5, Firefox, the sky, etc.), but no one stops to think about MPEG-LAs past actions.
As far as I can tell, MPEG-LA has never gone after anyone, especially open-source projects. x264 is a perfect target; they are massively guilty of patent infringement, but MPEG-LA doesn't do anything about them.
Yes, it's bad that MPEG-LA won't come out and say, "x264, we promise not to sue you." But, the absurd amounts of fear surrounding h.264 are not helping anyone.
FUD Posted Aug 31, 2010 17:17 UTC (Tue) by JoeBuck (subscriber, #2330) [Link] On what grounds would they sue a free software project that operates in a jurisdiction that does not recognize software patents? And if they did, how would they hope to even earn back their legal costs?Patent suits can cost millions, so they only pay when there are millions to collect and a reasonable chance of collecting. Mozilla's revenues are in the tens of millions, so they are vulnerable.
FUD Posted Aug 31, 2010 22:29 UTC (Tue) by drag (subscriber, #31333) [Link]
> On what grounds would they sue a free software project that operates in a jurisdiction that does not recognize software patents? And if they did, how would they hope to even earn back their legal costs?
They would not go after the people operating in the patent-free jurisdiction... they would go after people using the software in patent-encrusted nations.
Which they do.
They make a shitload of money from people using x264: far more then the people that wrote the software will ever will see.
I thought the days of 'It's open source, but not for commercial purposes' were far far behind us. It's amazing that people still exist that are so backwards that they would find it acceptable that licensing terms would demand that users and developers must never seek a profit.
x264 developers reside in the USA Posted Sep 1, 2010 1:40 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
> On what grounds would they sue a free software project that operates in a
> jurisdiction that does not recognize software patents?
The two main developers of x264, Loren Merritt and Jason Garrett-Glaser, reside in the USA; their company is there as well.
x264 developers reside in the USA Posted Sep 1, 2010 3:41 UTC (Wed) by coriordan (guest, #7544) [Link] They're probably safe because there's little commercial value in suing them, and their work is actually quite welcome since their work gives commercially significant entities the opportunity to become infringers. They're safe, but Red Hat probably wouldn't be safe to distribute their software. (I'm not sure how to check what packages are in RH's GNU/Linux distro. I just looked at this CentOS package list, but I don't know enough to know if that's a complete list, or a base install, or whatever. I don't see ffmpeg, or mplayer, or x264.)
multimedia in Fedora / Red Hat Posted Sep 1, 2010 11:58 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
Neither Red Hat nor Fedora distribute FFmpeg or multimedia software that builds upon it. They do not even ship MP3 decoders. Unfortunately this means that they do not cater to the multimedia needs of their users, which are forced to go with external package sources.
multimedia in Fedora / Red Hat Posted Sep 4, 2010 18:33 UTC (Sat) by luya (subscriber, #50741) [Link]
" Neither Red Hat nor Fedora distribute FFmpeg or multimedia software that builds upon it. They do not even ship MP3 decoders"
They do not ship them because of US patent laws. The needs of their users can be in free and open source format of media such as webm, theora, vorbis to name a few. See: http://fedoraproject.org/wiki/ForbiddenItems
External repository is called RPMFusion which provides package not included on either Fedora and Red Hat repository. Another alternative is to get mp3 and other patented codec from Fluendo.
multimedia in Fedora / Red Hat Posted Sep 7, 2010 18:04 UTC (Tue) by DonDiego (subscriber, #24141) [Link]
> Another alternative is to get mp3 and other patented codec from Fluendo.
This is not an alternative. Fluendo's codecs are proprietary, non-free crapware. Also, they are slower and buggier than FFmpeg. If you call this an alternative, you might as well go all the way and suggest installing Windows.
multimedia in Fedora / Red Hat Posted Sep 9, 2010 14:45 UTC (Thu) by paulj (subscriber, #341) [Link]
Google seem to manage to ship FFmpeg, while being an MPEG-LA licensee, presumably Fraunhoffer too. If that's within the licence of FFmpeg, and if you must pay those tithes (e.g. you must distribute MPx decoders), then why not do it to cover a free software implementation, rather than a proprietary one?
x264 developers reside in the USA Posted Sep 1, 2010 5:34 UTC (Wed) by sfeam (subscriber, #2841) [Link] If you mean Avail Media, note that it is an MPEGLA licensee according to this page.
x264 company resides in the USA Posted Sep 1, 2010 11:40 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
No, I'm talking about x264 LLC.
Patents relate to a territorial target market irrespective of legal domicile Posted Sep 1, 2010 4:30 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link] On what grounds would they sue a free software project that operates in a jurisdiction that does not recognize software patents? That's a wrong question to ask. Patents are a time-limited monopoly within a territorial target market. You can apply for them from anywhere else, and you can infringe them from anywhere else. Plus, as someone already explained here, there's no need to go after the original developer. Targeting distributors and/or commercial users in the local market would also work. That is actually the more usual thing to happen, although there are exceptions, such as Oracle going directly after Google instead of individual Android phone makers. By the way, what do you mean by "a jurisdiction that does not recognize software patents?" They exist in the US, Europe, Asia and Australia, as shown by the examples listed here (international equivalents of patents currently enforced in court by Apple, Oracle and Paul Allen). In New Zealand, the jury is still out and I'm not nearly as optimistic as some others are.
FUD Posted Aug 31, 2010 17:19 UTC (Tue) by vonbrand (subscriber, #4458) [Link] Just think SCOX (neé Caldera), Oracle's moves with stuff that used to belong to Sun. Sure, there is not (too) much to be afraid of now, but that can change radically on short notice. And the ripping out of GIF shows that the switch can be painful...
FUD Posted Aug 31, 2010 17:22 UTC (Tue) by Trelane (subscriber, #56877) [Link]
"As far as I can tell, MPEG-LA has never gone after anyone, especially open-source projects."
http://www.twice.com/article/257658-Audiovox_Disputes_MPE...
what's more, in January, they started suing via a company they own:
MPEG LA targets Posted Sep 1, 2010 12:11 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
All the examples you give are commercial entities, where the MPEG LA is vigilant and diligent about collecting their protection money. However, the MPEG LA has never sued an open source project.
MPEG LA targets Posted Sep 1, 2010 14:39 UTC (Wed) by Trelane (subscriber, #56877) [Link]
1) If you notice the quoted text, open source wasn't the only thing being discussed.
2) This will be of little comfort to commercial entities (e.g. GOOG) 3) Yet. Hooray for patent uncertainty!
MPEG LA targets Posted Sep 1, 2010 15:54 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
> 1) If you notice the quoted text, open source wasn't the only thing being
> discussed.
I neither meant nor implied only FOSS was being discussed, please reread my comment again.
> 2) This will be of little comfort to commercial entities (e.g. GOOG)
Google is an MPEG LA licensee for the H.264/AVC patent pool, they are not in need of comfort.
> 3) Yet. Hooray for patent uncertainty!
Sad as it is, there is no such thing as certainty with software patents. And as much as it hurts to say so, the MPEG LA is in the business of giving its clients patent certainty...
MPEG LA targets Posted Sep 1, 2010 16:24 UTC (Wed) by Trelane (subscriber, #56877) [Link] I neither meant nor implied only FOSS was being discussed, please reread my comment again.Fair 'nuff, I suppose. My point solely addressed the assertion that MPEG LA hadn't sued anybody. Hooray for implicit limitations. Google is an MPEG LA licensee for the H.264/AVC patent pool, they are not in need of comfort. For that one yes, and for the specific software and uses for which they paid proection money, not others ways in which they may be found to infringe. In addition, not all companies are GOOG nor are all downstream users of "Free" and Open Source software (quotes because the freedom of a software that is known by its developers to infringe on patents is IMHO highly dubious) are MPEG LA licensees of the relevant licenses. And as much as it hurts to say so, the MPEG LA is in the business of giving its clients patent certainty...Well, in as much as "we're not going to be sued by MPEG LA" is increasing the certainty. It's not the same as "we're not going to be sued for patent infringement". Rather like paying "protection money" helps ensure that your "protectors" won't be the ones to bust up your store and employees (but is no guarantee that nobody else will). Actually, perhaps MPEG LA licensing is even less of a guarantee than such "protection" schemes offer, because they defend their turf....
MPEG LA targets Posted Sep 1, 2010 22:35 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
> > Google is an MPEG LA licensee for the H.264/AVC patent pool, they are not in need of comfort.
> For that one yes, and for the specific software and uses for which they paid proection money, not others ways in which they may be found to infringe.
Where do you get this idea from? The MPEG LA licenses are not in any way specific to software. In what other ways do you say Google could be infringing and not be covered by MPEG LA patents?
> (quotes because the freedom of a software that is known by its developers to infringe on patents is IMHO highly dubious)
So the Linux kernel is not free? Is Samba not free? VLC player? Firefox? You can continue the list indefinitely I'm afraid.
There is of course software where the authors do not know it infringes on patents, but that will likely change for any nontrivial software if the authors started looking.
So is software free while the authors live in blissful ignorance, but no longer free once they start looking or somebody informs them of the fact that a patent is infringed?
> > And as much as it hurts to say so, the MPEG LA is in the business of giving its clients patent certainty...
> Well, in as much as "we're not going to be sued by MPEG LA" is increasing the certainty.
Of course it increases the certainty. The MPEG LA patent pools cover a lot of companies and a lot of patents. You seem to think that only absolute and complete certainty/protection is worth paying for and anything less is not worth a single dime. This is not true and obviously the MPEG LA licensees disagree. Getting a license was worth money for them and they did not get snake oil.
Seat belts do not give you absolute certainty/protection in car crashes. Nonetheless I'm sure that you will agree it's wise to use them.
I dislike software patents at least as much as the next guy. The system must be changed, but pretending it does not exist or claiming that the people forced to work inside it are all complete fools does not help anyone.
MPEG LA targets Posted Sep 1, 2010 23:32 UTC (Wed) by Trelane (subscriber, #56877) [Link]
Whatever, dude. Clearly, continued discussion with you is pointless. Perhaps we can continue this in its proper place--over a beer. Until then, whatever.
MPEG LA beer Posted Sep 2, 2010 20:19 UTC (Thu) by DonDiego (subscriber, #24141) [Link]
> Whatever, dude. Clearly, continued discussion with you is pointless.
> Perhaps we can continue this in its proper place--over a beer. Until then, whatever.
Gladly! :)
Do you attend, FOSDEM, LinuxTag or similar events in central Europe?
MPEG LA beer Posted Sep 2, 2010 20:21 UTC (Thu) by Trelane (subscriber, #56877) [Link]
I hope to someday, but I don't yet. Stupid obligations and responsibility and not being a full-time linux hacker. :(
Beers galore Posted Sep 8, 2010 19:37 UTC (Wed) by man_ls (subscriber, #15091) [Link] We need a lot of beers in a lot of places (and not only for its delicious foam and nutritive properties). Luckily you are mostly agreeing on the big picture.
MPEG LA targets Posted Sep 2, 2010 2:50 UTC (Thu) by cmccabe (subscriber, #60281) [Link]
> Sad as it is, there is no such thing as certainty with software patents.
Yep.
> And as much as it hurts to say so, the MPEG LA is in the business of
Come on, this has been discussed to death.
Suppose I manage to get a patent that bears on x264, or video decoding in general. Then I can form MPEG-LA 2.0 and start suing all the same people that MPEG-LA 1.0 could sue.
Paying MPEG-LA doesn't give you certainty of anything, except that your bank balance will be lower afterwards. In a world where "one-click ordering" or putting an LCD screen next to an e-ink screen can be patented, it's not unreasonable to suppose that there is an MPEG-LA 2.0 out there already.
MPEG LA targets Posted Sep 2, 2010 3:03 UTC (Thu) by cmccabe (subscriber, #60281) [Link]
I almost forgot to mention the most important part. MPEG-LA 1.0 can't sue 2.0. They are both "non-practicing entities", so they cannot be sued for patent infringement.
So there is no way that the MPEG-LA can defend its turf through the legal system. I guess it could try to coerce its "customers" to attack MPEG-LA 2.0 in court, but that's not likely to benefit the customers / victims since they'll end up spending millions more.
MPEG LA targets Posted Sep 2, 2010 20:22 UTC (Thu) by DonDiego (subscriber, #24141) [Link]
> Suppose I manage to get a patent that bears on x264, or video decoding in
> general. Then I can form MPEG-LA 2.0 and start suing all the same people > that MPEG-LA 1.0 could sue.
In theory yes, in practice it's not so easy to get new patents on H.264 and it's only the MPEG LA going around extorting people for protection money.
Besides, a seat belt only protects you certain types of car accidents. Would you therefore not recommend wearing one?
MPEG LA targets Posted Sep 3, 2010 12:10 UTC (Fri) by pboddie (guest, #50784) [Link] Suppose I manage to get a patent that bears on x264, or video decoding in general. Then I can form MPEG-LA 2.0 and start suing all the same people that MPEG-LA 1.0 could sue. I thought MPEG-LA 2.0 existed already.
Who's shipping it now? Posted Aug 31, 2010 17:28 UTC (Tue) by coriordan (guest, #7544) [Link]
Do commercial GNU/Linux distros (Red Hat, SuSE, Mandriva, (Ubuntu?)) ship h.264 support in the default install? Are they paying licence fees?
I think that's how we'll see whether MPEG-LA is ignoring infringement or if they're just waiting until the right moment.
Who's shipping it now? Posted Aug 31, 2010 17:51 UTC (Tue) by dowdle (subscriber, #659) [Link]
The only mainstream distro that ships with h.264 support that I know of is Ubuntu... because they have paid for it. I'm not sure exactly how that works or how much they are paying.
Red Hat and Fedora do NOT ship with h.264 support... only webm and ogg and a few other non-patent encumbered codecs... which is a big reason several third-party repos are popular.
I'm not sure about SUSE.
I'm not sure about Debian. h.264 is probably available in their non-free repos but I don't think it ships as part of the install media.
I guess the better answer/question is... anyone who ships with ffmpeg and/or mplayer/mencoder... unless they've stripped out the patented stuff... ships with h.264 support.
Who's shipping it now? Posted Aug 31, 2010 17:58 UTC (Tue) by dowdle (subscriber, #659) [Link]
Hmmm, or am I confusing that with mp3 support in Ubuntu?
Who's shipping it now? Posted Aug 31, 2010 18:09 UTC (Tue) by jspaleta (subscriber, #50639) [Link]
I think if you look closely the license its only for certain OEM partner pre-installs. The h.264 license situation calls for a level of situational awareness that is I doubt most people can achieve without the aid of cutting edge pharmaceutical enhancements to extend their perception into hidden dimensions of reality.
http://www.h-online.com/open/news/item/Canonical-clarifie... :
"So the rule of thumb is that an arbitrary Ubuntu system does not have a H.264 licence via Canonical, unless it's an OEM system which specifically lists the H.264 licence in its documentation or marketing materials."
Can anyone point to an existing Ubuntu pre-install OEM which goes out of its way to list H.264 license compliance in its marketing or docs?
-jef
Who's shipping it now? Posted Sep 1, 2010 12:50 UTC (Wed) by dgm (subscriber, #49227) [Link]
I could find nothing in the materials that came with my Dell Latitude, nor in the Dell's site.
Who's shipping it now? Posted Aug 31, 2010 18:36 UTC (Tue) by foom (subscriber, #14868) [Link] Debian ships all decoders that are part of ffmpeg upstream in the main repository.From /usr/share/doc/libavcodec52/README.Debian.gz: Currently the following video encoders are disabled in the ffmpeg package: H263, H264, MPEG2 video, MPEG4 and MS-MPEG4. No *decoders* are disabled in any the ffmpeg package!
Who's shipping it now? Posted Aug 31, 2010 20:01 UTC (Tue) by coriordan (guest, #7544) [Link]
Debian (my distro of choice) is safer than most distros since there's no wealthy organisation to sue, that's why it's probably more relevant to look at "commercial" distros.
Debian ships all of FFmpeg Posted Sep 1, 2010 1:45 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
Debian no longer strips out any encoders. Uncrippled packages have been uploaded to unstable a short while back before the freeze, so they will appear in the next stable release.
Debian ships all of FFmpeg Posted Sep 1, 2010 7:34 UTC (Wed) by patrick_g (subscriber, #44470) [Link] Are you sure about that ? No need to add the Debian Multimedia repository in the sources.list ?
Debian ships all of FFmpeg Posted Sep 1, 2010 12:01 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
Who's shipping it now? - Ubuntu is Posted Sep 1, 2010 12:08 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
Ubuntu has FFmpeg (uncrippled) as well as x264 in universe. They leave them out of the install DVDs that they send you for free. Does anybody care about those DVDs? I don't think so...
FUD Posted Aug 31, 2010 17:34 UTC (Tue) by zlynx (subscriber, #2285) [Link]
I am not a lawyer, but as far as I know, source code can never violate a patent.
Only the "machine", which is the combination of a computer executing a program, can violate the patent.
I'm not sure if a lawyer could go after the source code for contributing to patent infringement -- possibly.
But distributing source code cannot possibly be any more illegal than distributing copies of the patent application documents. Patents are intended to promote the spread of the technical arts, after all, and the patent documents are supposed to describe the complete machine in a reproducible way. So really, if algorithm patents are allowed at all, they *should* contain a complete source code sample implementation.
FUD Posted Aug 31, 2010 17:57 UTC (Tue) by dowdle (subscriber, #659) [Link]
During the time that "DVD Jon" was in trouble... weren't the litigators sending out cease and desist letters to sites that had posted the source for his DVD decrypting program? There were also tee-shirts with the code on it... and audio recordings of the source code being read and sung. This was in defiance to the litigators who said that such source was illegal.
I don't recall exactly what happened with regards to the whole, "is source code a violation?" part of it but I believe in the eyes of the law (at least in certain countries) it is.
FUD Posted Aug 31, 2010 20:57 UTC (Tue) by cortana (subscriber, #24596) [Link]
"DVD Jon" ran afoul of the DMCA's anti-circumvention laws. Not patent infringement.
FUD Posted Aug 31, 2010 18:01 UTC (Tue) by coriordan (guest, #7544) [Link] > source code can never violate a patent. My instincts tell me this is wrong, but I noticed an interesting thing when looking for proof. In the USA, the relevant part of the Patent Act § 271, and in Justice Ginsburg's ruling in Microsoft v. AT&T (2006, USA) that "Because no physical object originating in the United States was combined with these computers, there was no violation of §271(f).". That case hinged on the word "component" in paragraph (f), so it doesn't necessarily change the main paragraph (a), but it would be interesting to look into what other types of infringement might not be applicable to software. > But distributing source code cannot possibly be any more illegal than distributing copies of the patent application documents. Distributing software is inducing infringement :-/ Distributing an MPEG patent won't cause anyone to enjoy a video.
FUD Posted Aug 31, 2010 18:08 UTC (Tue) by linuxrocks123 (guest, #34648) [Link]
I have two things to say about this. First, the case with DVD Jon was never fully resolved, but just kind of dropped, so it has little value. Second, the DVD Jon case was about copyright infringement, not patents, so it is not relevant here in any case.
Distributing software could, arguably, be patent infringement. However, there is an independent experimentation defense to patent infringement which very well may apply. Also, you could say that you are distributing the software to help other independent experimenters or for use by people who do have a patent license. As far as I can tell, no one has ever been sued for distributing source code, so we really just don't know.
---linuxrocks123
FUD Posted Sep 1, 2010 17:03 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
> I have two things to say about this. First, the case with DVD Jon was
> never fully resolved, but just kind of dropped, so it has little value.
Where did you get the idea? DVD Jon won fair and square, then the appeal was laughed out of court...
FUD Posted Sep 20, 2010 23:53 UTC (Mon) by linuxrocks123 (guest, #34648) [Link]
My apologies, this was conflating a few issues. You are correct that DVD Jon won fair and square in Norway. What I was thinking of were the subsequent cases in the U.S. against various people not Jon. I seem to recall that these cases were, ultimately, dropped by the MPAA, although I don't care to look them up right now. So, there's little legal precedent in the U.S. for this that goes our way.
---linuxrocks123
Software can be infringing material and software publishing an infringement Posted Sep 1, 2010 4:36 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link]
The claim that "source code can never violate a patent" is difficult to makie under US patent law. In my jurisdiction (Germany), distribution or commercial use would be required to make it a patent infringement. Publishing software on the Internet (including the upload to a source code repository that's accessible by third parties) is clearly an infringement everywhere.
Oracle also goes after Google directly instead of targeting the makers of Android-based phones, although Google just makes the code available.
The other point you make (patents should be required to "contain a complete source code sample implementation") is not a matter of infringement. This is a question of patentability requirements. Disclosure (alongside novelty and nonobviousness and others depending on the jurisdiction) is one of them. I, too, believe that disclosure requirements for software patents should be much higher. The term "algorithm patent" is however tricky because there are many patents that effectively monopolize algorithms but are filed for as "apparatus" claims.
Software can be infringing material and software publishing an infringement Posted Sep 2, 2010 1:56 UTC (Thu) by rahvin (subscriber, #16953) [Link]
Are you kidding? Source code falls under free speech. We had laws in the 90's about exporting crypto that made is nearly impossible for companies like PGP to do business. To export the crypto they simply exported the source code only because the source code was protected under free speech laws. (and to be absolutely certain the published the source code in paper format as a book) The only time Source Code would even have the possibility for litigation would be copyright or possibly the DMCA (but a DMCA violation would need to survive a constitutional free speech challenge which IMO it wouldn't survive).
Source code is free speech, until it's complied and used it's nothing but free speech. Any attempt to regulate it would very likely fail a constitutional challenge. This is the reason you won't find case law on it because no on has at present been stupid enough to try to challenge the free speech protections source code enjoys.
I might add that the Supreme court in the US struck down campaign finance laws because they didn't believe the lower courts upholding of the ban using the idea that "commercial" speech wasn't protected. In other words, all forms of speech, including commercial are protected regardless of the laws congress passes in the interim. Patent litigation against source code wouldn't make it past the first dismissal filing because without compilation the code is nothing but speech.
Now if the code was compiled and executable, that's another game all together.
Software can be infringing material and software publishing an infringement Posted Sep 2, 2010 2:50 UTC (Thu) by foom (subscriber, #14868) [Link]
Hm.... Is there really so much of a difference between speech in the Python language and speech in the x86 binary language? Both are executable, both are understandable by [some] humans.
Software can be infringing material and software publishing an infringement Posted Sep 2, 2010 4:07 UTC (Thu) by FlorianMueller (subscriber, #32048) [Link]
Oracle is suing Google over patent infringement (and inducement to infringement) through providing source code, which others then compile and distribute.
Like foom just commented, there's no particular reason to consider source code more of a free speech issue than binary code.
FUD Posted Aug 31, 2010 18:33 UTC (Tue) by robert_s (subscriber, #42402) [Link]
MPEG-LA exist for nothing more than "going after people" and collecting fees from people who pay up.
MPEG-LA Posted Aug 31, 2010 22:03 UTC (Tue) by tialaramex (subscriber, #21167) [Link]
actually that's a relatively minor role, as with most such organisations.
The main role is what we would in happier times call "keeping honest people honest". All the companies for whom these patents are just a cost of doing business have a single entity that can bill them, make decisions about who to audit, arrange cross licensing deals to keep things simple. AFAIK MPEG-LA is the same from MPEG-1 which was developed when dinosaurs roamed the Earth. Obviously to collect these fees easily it helps the MPEG-LA to have some reputation for being litigious, but that doesn't make litigation their main role.
Among patent believers (or even companies that just reluctantly accept them) MPEG-LA makes life a lot easier than having say two dozen patent holders wanting separate license deals, for which you have to provide different compliance data. Having dealt with buying information (ie copyright, and sometimes dubious database rights, not patents) I can tell you that I'd even pay slightly more to get everything from one place, just because the constant contract renegotiations are a huge waste of time. No two suppliers ever use the same terms, it's almost as though they're psychic and can tell which terms you already have numbers for, just to make you do more work.
The MPEG-LA is probably the best test of anti-patent feeling, because these aren't pure trolls, several of the member organisations spent a pile of money on developing fundamental video encoding techniques. Only if we're sure that software patents are in principle wrong (as I am) can we say with confidence that it's wrong for them to collect license fees on implementation of the standard -- a milder stance would be that MPEG should have insisted on free patent grants rather than RAND terms for development, but that would probably have just killed the process.
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 4:47 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link] MPEG LA isn't a judge who allows software patents nor a politician deciding to condone the grant of software patents instead of intervening through new and restrictive legislation. Under the existing circumstances (for as much as I'd like software to be clearly excluded from the scope of patentable subject matter), MPEG is part of a commercial solution, not part of the (legal/political) problem. It's just that everyone can relate to the topic of digital video. So a lot of people hate the notion that something so fundamental should be so very patent-encumbered. But the companies behind MPEG LA would have filed for those patents anyway, and if an MPEG-like one-stop licensing solution with so very reasonable terms from a commercial point of view were available in all fields of software technology, this industry as a whole would clearly benefit from it. There are strategic reasons on the part of companies like Google (which actually owes its success in no small part to patent protection and started out with a patent before a complete product and is a pro-software-patent company) why they don't want web video to be patent-encumbered. But in terms of where the really serious software patent problems are, MPEG is the wrong place to search. Exclusionary strategic use of patents is the real issue; something like MPEG is a solution under the circumstances and Google and others self-servingly exaggerate the problem it represents. I urge everyone not to be misled by diversionary tactics by those who seek to distract from the real patent problems. For an example, the way IBM uses its mainframe-related patents are a hugely bigger problem, in strategic and economic terms (bigger than even Google as a whole), than the MPEG LA kind of licensing business..
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 9:54 UTC (Wed) by roc (subscriber, #30627) [Link]
If MPEG-LA didn't exist, the H.264 situation would be so obviously unworkable that a patent-unencumbered solution would take over the market. So in that sense, MPEG-LA enables patented formats.
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 10:02 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link]
This is an interesting line of thought, but it all depends on whether or not the patents held by the contributors to the MPEG LA pool are such a thicket that there's no designing around. I know that many proponents of VP8, among them some really competent ones, say that the format is patent-unencumbered. But there also those who say that all codecs are based on only two or three fundamental ideas. There are only two ways in which we can all find out: either there's no assertion of patents against VP8 for a year or more, in which case it's presumably safe, or there will be infringement claims, in which case we'd have to see whether those are legally strong.
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 10:29 UTC (Wed) by rahulsundaram (subscriber, #21946) [Link]
"I know that many proponents of VP8, among them some really competent ones, say that the format is patent-unencumbered"
Reference needed. So far, it all seems to be coming from H.264 proponents.
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 13:18 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link] Reference needed. So far, it all seems to be coming from H.264 proponents. There's no analysis out there yet by people who don't favor one model (open source or patents) over the other. You'll find FOSS advocates supporting Google (I particularly recommend reading this analysis by Carlo Daffara) and claims to the contrary such as this one by an x.264 developer, which was endorsed by Steve Jobs (who emailed to a user a link to that analysis). Then there's MPEG LA's confirmation that it contemplates creating a patent pool for VP8. It doesn't help to dismiss those kinds of claims as FUD. In my opinion, the burden of proof is on Google as the late entrant. H.264 has no patent problems so far despite its ubiquity, which is kind of empirical evidence. Plus there are H.264 adopters who provide software with an intellectual property guarantee. But if Google is so sure that VP8 has no patent problems, why doesn't it (i) publish an analysis as to why VP8 doesn't infringe on any MPEG patents and (ii) provide indemnification? My assumption is: they aren't as sure they claim.
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 13:58 UTC (Wed) by rahulsundaram (subscriber, #21946) [Link]
Your answer is not related to the question I asked. Where did VP8 proponents claim that it is patent encumbered? Repeating yourself over and over again doesn't substantiate your claim.
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 14:02 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link] The problem is that you misread, not that I repeat. I just tried to explain this but you made a mistake in the first place. Your original question was this: "I know that many proponents of VP8, among them some really competent ones, say that the format is patent-unencumbered" So read it again. I said that VP8 proponents claim it's patent-unencumbered. When you asked your question, I thought you wanted to see references for the different claims. I thought you had just quoted the wrong part, but now it's clear you misread it.
Software patentability is the problem but licensing is a commercial solution Posted Sep 2, 2010 10:29 UTC (Thu) by roc (subscriber, #30627) [Link]
If Google thought VP8 might have patent problems, they probably wouldn't have blown $100M buying it.
Software patentability is the problem but licensing is a commercial solution Posted Sep 2, 2010 12:30 UTC (Thu) by FlorianMueller (subscriber, #32048) [Link]
It's not that simple. The amount was even greater (closer to $200M), but that's still different from Google betting the house on something. The price they paid is a fraction of the strategic value of controlling its own video format, so the price all by itself is too low to reflect an assumption of near-certainty on Google's part. Actually, it's a price that could simply be probabilistic, such as "worth it if there's only a 15% chance of success" (you could replace the 15% with any other number you're more comfortable with).
A probabilistic approach would be based not only on the assessment of a risk of patent infringement but even if one were pretty certain, Google could gamble with the determination of others to enforce it. They could hope that before others take action VP8 gets enough traction in the market, possibly even W3C adoption in a best-case scenario for Google.
I don't mean to say that the low price is necessarily probabilistic. It's low enough that it could be; they might also simply have gotten a bargain deal.
Software patentability is the problem but licensing is a commercial solution Posted Sep 2, 2010 20:27 UTC (Thu) by DonDiego (subscriber, #24141) [Link]
> If Google thought VP8 might have patent problems, they probably wouldn't have blown $100M buying it.
You assume that the only or main reason for Google to buy On2 was to get video coding technology without patent issues. This is pure speculation. On2 has other values apart from that.
Software patentability is the problem but licensing is a commercial solution Posted Sep 2, 2010 10:35 UTC (Thu) by roc (subscriber, #30627) [Link]
Those claims are FUD as long as the claimants refuse to name a single patent they believe VP8 infringes.
Software patentability is the problem but licensing is a commercial solution Posted Sep 2, 2010 12:33 UTC (Thu) by FlorianMueller (subscriber, #32048) [Link]
Not exactly. I agree to the extent that any such claims might be FUD until proven to be well-founded. But the fact that the owners of the potentially relevant patents haven't come up with specific claims yet in public doesn't mean much. They might have done so in private; or they might not even have done so in private but simply take more time before going on the offensive.
I don't like the fact that patent holders don't have to make every infringement assertion and every license deal public, but it makes sense not to have such a requirement in place because patents are in almost every high-tech product so patent-related transparency would result in total business transparency that just wouldn't work. So patent holders and their licensees can do backroom deals without ever even talking about them; or patent holders can contact alleged "infringers" and present them with claim charts only under a non-disclosure agreement.
The burden of proof on Google Posted Sep 1, 2010 13:21 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link]
An addition to what I just wrote about Google having the burden of proof. Let's just imagine this scenario: someone comes up with a new search engine and says it's patent-free and encourages its widespread use also by businesses in ways where they might be liable for patent infringement. Google and/or organizations in Google's camp say that the thing infringes on Google patents. Obviously, since Google is the long-time market leader in that field, one would have to take that seriously, and the new entrant would have to (i) document that there's no problem and (ii) provide indemnification before customers would take any chances.
Software patentability is the problem but licensing is a commercial solution Posted Sep 4, 2010 23:56 UTC (Sat) by gmaxwell (subscriber, #30048) [Link]
"there also those who say that all codecs are based on only two or three fundamental ideas"
To the extent that this statement is true it's also an argument for the patent-freeness of codecs like VP8. It's true that all current popular video codecs share common set of "two or three fundamental ideas", e.g. they are all block oriented transform codecs. But codecs of this specific lineage have existed for so long that those fundamentals must have lapsed if they were ever patented.
Of course details matter and the way those fundamental concepts are used may result in patentability but once you've admitted details you must also admit that all the current codecs differ significantly at the level of detail required for determining patent exposure.
Please don't spread this FUD... and if you insist on spreading it, please have the courage to assign your own name to it rather than casting it towards a undefined (and perhaps non-existing) 'those'.
Software patentability is the problem but licensing is a commercial solution Posted Sep 1, 2010 22:17 UTC (Wed) by DonDiego (subscriber, #24141) [Link]
> If MPEG-LA didn't exist, the H.264 situation would be so obviously
> unworkable that a patent-unencumbered solution would take over the market. > So in that sense, MPEG-LA enables patented formats.
There are other formats that did not have a patent pool behind it and still succeeded in the market, so I'm not sure your argument holds water. The first and foremost example is MP3.
Software patentability is the problem but licensing is a commercial solution Posted Sep 2, 2010 10:27 UTC (Thu) by roc (subscriber, #30627) [Link]
MPEG-LA says that dozens of companies hold hundreds of patents essential for H.264. That's a lot worse than MP3.
Software patentability is the problem but licensing is a commercial solution Posted Sep 4, 2010 23:49 UTC (Sat) by gmaxwell (subscriber, #30048) [Link]
When Mpeg LA asked the DOJ for the permission to exist for MPEG2 they specifically cited the thicket and promised that the pool would be strictly limited to patents which were essential and irreplaceable for the implementation. The DOJ's analysis was conditional on these facts. (this is one reason why any particular real implementation may still violate patents outside of the pool a technique might be customary or even important, but if it's not essential it can't be in the pool)
If it were the case that there weren't a thicket and that the license bundled inessential patents then MPEG LA and the patent holding companies would be breaking a multitude of anti-trust laws.
Basically highly patented technology suffers from a substantial disadvantage in the market by its very nature. Pooling organizations enable a kind of collusion which reduces this disadvantage.
Software patentability is the problem but licensing is a commercial solution Posted Sep 7, 2010 18:08 UTC (Tue) by DonDiego (subscriber, #24141) [Link]
Interesting. Can you provide some background material on the creation of the MPEG LA?
FUD Posted Sep 1, 2010 11:31 UTC (Wed) by __alex (subscriber, #38036) [Link]
x264 is not infringing because there are exclusions in the licensing for development purposes. Only individual deployments of x264 require licenses.
Simon Phipps and patents Posted Sep 1, 2010 4:53 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link]
I like Simon's current resistance to software patents but it should be mentioned that he spent a large part of his professional life at Sun, and Sun filed for patents quite aggressively and used them (such as against Microsoft). As James Gosling explained, they only became very active on the patenting front after a patent holdup by IBM, but that's what they did and they never pushed for the abolition of software patents. Jonathan Schwartz did some very pro-software-patent blog postings (not the only area in which he was #fail).
Simon Phipps and patents Posted Sep 1, 2010 12:21 UTC (Wed) by alankila (subscriber, #47141) [Link]
Wasn't the Microsoft attack a trademark dispute about Java, or are you talking about some different thing? If so, that wasn't a patent suit, but a contractual violation about the correct use of the Java trademark.
Simon Phipps and patents Posted Sep 1, 2010 13:09 UTC (Wed) by FlorianMueller (subscriber, #32048) [Link] The announcement at the time broke up the $1.6 billion settlement into $900 million for antitrust issues and $700 million for patent issues. In addition, an upfront royalty payment of $350 million was announced. That was more of a contractual thing.
|
|
||||||||||||