Slashdot Mirror
MPEG LA Attempts To Start VP8 Patent Pool
Confirming speculation from last year, an anonymous reader tips news that MPEG LA has posted a request for information about establishing a patent pool for the VP8 video codec.
"In order to participate in the creation of, and determine licensing terms for, a joint VP8 patent license, any party that believes it has patents that are essential to the VP8 video codec specification is invited to submit them for a determination of their essentiality by MPEG LA’s patent evaluators. At least one essential patent is necessary to participate in the process, and initial submissions should be made by March 18, 2011. Although only issued patents will be included in the license, in order to participate in the license development process, patent applications with claims that their owners believe are essential to the specification and likely to issue in a patent also may be submitted."
Ultimately evil, this will at least provide a list of patents that threaten VP8.
There have been previous threats about Theora, but nothing happened. This could be FUD bluff too.
MPEG LA has over 1,000 patents which it says are needed for an implementation of MPEG H.264.
Most current efforts around software patents talk about quality and bad patents, but none of those efforts will make a dent on a thicket of 1,000 patents. It's unlikely they're all obvious or that prior art exists for them.
* http://en.swpat.org/wiki/MPEG_LA
* http://en.swpat.org/wiki/WebM_and_VP8
* http://en.swpat.org/wiki/Audio-video_patents
There was a time when the problem was killer patents - RSA, public key encryption, gif - but today the problem is always thickets. Raising quality won't solve the problem, we need to clearly exclude software from patentability.
Expert in software patents or patent law? Contribute to the ESP wiki!
If you think you have a patent to use against Vp8, signing it over to MPEG-LA is the last thing you should do. Look at their other pools, the few good patents are swamped by the dross. The earnings from your patents get shared with troll patents of no value.
Even the CEO of MPEG-LA HAS A SEPARATE patent pool all for himself largely undiluted? His MobileMedia Ideas LLC, is like a patent pool, only he's bought patents and controls the pool.
If he won't throw those patents into the diluted pool, neither should you.
http://www.osnews.com/story/23258/MPEG-LA-owned_Patent_Troll_Sues_Smartphone_Makers
Businesses don't care about open source. They care about patent liability. Using a codec that can get you sued isn't in their best interests.
Nice shiny-looking coded you've got there. It would be a pity of something bad were to happen to it. You know that you won't have to worry about any of these legal threats if you just license h.264, and it's just a small fee. For a short time, we've even waived the fee, just for you!
I know what MPLA is doing is technically legal, but spiritually it's corrupt. The patent system is broken, from many directions. For the other way around, my employer is routinely trolled by NPEs.
The living have better things to do than to continue hating the dead.
you either do not write codecs
or
you do not expect to be caught writing a H.264 codec.
MPEG-LA is the patent troll/layer firm that wants to collect $ for Software Patents on mathematical equations that compresses bits of information that happens to be a video. If it can scare/threaten/sue people into buying a license to create/use/distribute a H.264 codec then they will have accomplish their goal.
This WebM patent pool is a FUD tactic similar to MS' patents on Linux. You are free to create a WebM codec as much as you are free to write a Linux derivative. not so with H.264.
Obama's legacy: (N)othing (S)ecure (A)nywhere and (T)error (S)imulation (A)dministration
And the judges will live in Texas, the IP wonderland of the USA.
Obama's legacy: (N)othing (S)ecure (A)nywhere and (T)error (S)imulation (A)dministration
Dunno if you've ever actually read any MPEG specifications (I have), but the "open" standards process is not the best deal on the planet. Remember 802.11 "Draft N" and it's 5-year lifespan? That was a result of the standards process being held hostage by every vendor wanting to include their own pet features and patents. Constructing a standard around all those corporate filibusters took far too long, and the resulting standard is bloated and insanely complex. MPEG-4 and related (H.264 such) are stupidly huge standards that support all sorts of arcane features that nobody will ever use except as a) a marketing slide, and b) a patent to toss into the ring.
As per your claim that "anyone" can contribute to the MPEG standards, have you looked at the list of people involved? Every one of them comes from a major vendor or "research" company (or their university proxies) and is paying stupid sums of money to join and attend (meets move all over the planet). The reality is that if I had an idea I wanted to propose for an MPEG standard, hell will freeze over long before I even get a hearing let alone add the feature. Google is far more likely to pay attention to their (codec-savvy) users' comments.
I'll take unencumbered codecs with a single primary party keeping control over the standard every time.
GStreamer - The only way to stream!
Patents today are massively abused. Originally developed for the purpose of encourage innovation. tnhey actually squelch it. Many innovations improve on a previous idea. Patents make it difficult for independant developers to improve on and rrefine an idea. Patents have been turned in to a way to squelch independant innovation and create monopoliies, and stagnant technologies which are difficult to improve on or use independantly. Patent holders often charge licence fees so high they preclude independant use or refinement, make the technology too costly to deploy and limit its use, or they do not licence the technlogy at all, gaining a monopoly on ideas.
Both copyright law and patent law badly need reform. Software patents need to be prohibited altogether. Other patents need to be required to be licenced to other persons, for a reasonable percentage of profits (if there is no revenue there would be no fee). Copyrighted works no longer under production or distribution will also be made public domain immediately and must comply with the same reasonable licencing requirements.
H.264 is "open" like the Microsoft Open Office XML standard is open. There are bear-traps that can prevent you from using it successfully. In the case of OOXML, one of those bear-traps is requiring your documents to interpret "Proprietary Microsoft Binary Blobs".
H.264, being patent protected, means that the patent holder reserve the right to say, "Go pound sand" when your five-year license expires.
WebM is controlled by Google.... but Google has effectively put its patents into the public domain so that ANYBODY can implement a WebM/VP8 video codec.
Since anyone can, will and have sued anyone and anything else over anything, it seems that you are proposing that noone ever user codecs?
H.264 is "open" like the Microsoft Open Office XML standard is open. There are bear-traps that can prevent you from using it successfully. In the case of OOXML, one of those bear-traps is requiring your documents to interpret "Proprietary Microsoft Binary Blobs".
I'd call your comparison quite pathetic. h.264 is actually a useful standard. It provides a very, very clear specification for its compression and decompression, unlike WebM. It is widely used with a huge range of hardware implementations. It gives much better compression than WebM will ever have. It was created with the intention to produce the best possible video compression. From a technical point of view it is clearly superior to WebM, and Google is right now trying to force a sub-standard video compression onto the whole world. The OOXML standard, on the other hand, is ten times the size of the h.264 standard. It is full of bugs, unclear, unspecified in many areas, there are no implementations now and there will likely never be implementations (since Microsoft doesn't have any products that match this standard), and all these huge disadvantages are totally intentional.
The answer is yes, I did find at least one instance, and according to the article it is possible to win such a suit. This charming lawsuit
Chamilia’s slander of title claims also failed. Allegations of patent infringement can be slander of title if they are false, reasonably calculated to cause harm, and result in special damages.
That's an excerpt from the article relevant to the topic.
Has [a claim of slander of title] been used in the past against a patent holder??
Have you asked Google? The first result is a Groklaw article from the SCO v. IBM era. It cites an article on FindLaw; the link there is broken, but I found it with Google. Slander of title is the malicious publication of false and disparaging words by which special damages were sustained, causing a plaintiff who owns the property disparaged to lose a sale. FindLaw's article cites Prosser and Keeton (apparently Prosser and Keeton on Torts) that patents or other intangible property may be the subject of such a claim.
You are free to create a WebM codec as much as you are free to write a Linux derivative
Actually, I'm not entirely sure about this. The patent license appears to only apply to:
the contents of this implementation of VP8
So, unless I'm misreading it, you are only granted a license to use the patents with libvpx and derivatives, not with independent implementations. The implementation in ffmpeg (which is faster and smaller) would not be covered. Maybe someone can persuade Google to clarify this.
I am TheRaven on Soylent News
No, he said he'd be very surprised if it didn't step on patents because it's similar to parts of H.264 that are patented. He didn't, however, cite any specific patents, nor the claims in them that are violated by specific algorithms in VP8. Until someone does, then it's FUD.
I am TheRaven on Soylent News
Document formats are what maintains Microsoft's Office monopoly. Need I say more?
You can lead a man with reason but you can't make him think.
I've done quite a lot of professional video streaming via website for large well known companies. They don't know/care/understand open source, they don't know/care/understand patent liability.
They want max bang for buck, where bang is video quality and buck is filesize/lack of streaming problems on a shitty connection.
That's still h264. It's often *stunning* in it's ability to compress quality video. WebM becomes defacto standard when it's a better tool than the alternative.
To compare this to png/jpg. They both have good uses, but jpg does better compression of photographic images, which is why your porn collection doesn't have any png's in it.
Quoting from wikipedia: AMD, ARM, and Broadcom have announced support for hardware acceleration of the WebM format.[31][32] Intel is also considering hardware-based acceleration for WebM in its Atom-based TV chips if the format gains popularity.[33] Qualcomm and Texas Instruments have announced support,[34][35] with native support coming to the TI OMAP processor.[36] Chip&Media have announced the fully hardware decoder for VP8 that can decode full HD resolution VP8 streams at 60 frames per second.[37]
No. VP8 already has better compression efficiency than h.264 and at the same time being on par in quality with h.264. From the technical point of view, WebM has the potential to be a lot better than H.264.
It takes a man to suffer ignorance and smile
Be yourself no matter what they say
Both codecs are free in different ways. WebM is gratis; free as in beer; it costs no money to implement and distribute, but not to contribute to. H.264 is libre; free as in speech; you must pay to distribute an implementation, but anyone can contribute to the spec.
No.
WebM has a frozen bitstream standard. You cannot contribute to this standard, because it is frozen. However, you are free to contribute to the support software: encoders, decoders, tools. And there is vast room for improvement of the encoders, plenty of work there to keep you busy.
H.264 also has a frozen bitstram standard. During the development phase, which is now over, you could have contributed to the development process, but only if you could afford the fees to attend the meeting. I am not sure exactly how much that cost, but I saw one posting here on Slashdot claiming it was on the order of $40,000 per person per meeting. In fairness, you could probably pitch your ideas for free to one of the large companies that was sending representatives (e.g. Microsoft, Apple) and possibly get your ideas in anyway.
H.265 is under development now. Go ahead and try to contribute some ideas to it. Please report back here on Slashdot for how well that works out for you.
So, the important thing here is: H.264 and WebM are both frozen standards. Both have free software projects implementing them, to which anyone can contribute. But only one of them is fully free. H.264 is controlled by MPEG-LA, so you can only distribute H.264 (or even use H.264) with their permission, on their terms. If they decide "no H.264 licenses will be offered to Linux users" then the Linux users will have no legal way to use it. (That's a silly, extreme example, and I have no reason to think they will do that. However, they really are asserting that the mere act of using a camera that records in H.264 makes you subject to their whims, which is intolerable.)
Like I said, I think the ability to contribute the codec vastly outweighs the ability to implement it for free.
I'm sorry, but H.264 is frozen too. You are simply mistaken on this point.
steveha
lf(1): it's like ls(1) but sorts filenames by extension, tersely
The worst part of the MPEG-LA style open process is that it works to maximize the number of patents that show up in any new specification. Rather than delivering the best technology, it delivers maximally patent encumbered technology that's good enough to meet the requirements of the project. Everyone wants a piece of the pie, and in particular, the large companies who have an interest in earlier specs and don't want to be locked out of the next one.
-Dave Haynie
I never understood this one. If anything, clever algorithms should be the ONLY software I am allowed to patent (one-click is not clever, GIF's compression was). These people worked hard to come up with a novel idea that they then want to make money from. What is wrong with that exactly? Why SHOULDN'T they be able to patent them? We let people patent novel ways of building a mousetrap. If my mousetrap is virtual, why is that any different?
Two big differences.
First, algorithms are not patentable in the United States, and a lot of these so-called "software patents" are actually effectively patenting the underlying algorithm, since there's no other reasonable way to implement the algorithm. Yet they still get away with it because patent clerks don't understand that the patent application someone is filing is effectively pre-empting all implementations of the algorithm, so they approve the patent. Related to this is the issue of obviousness. Patent clerks are not able to determine with software what is and what isn't obvious. Using your analogy, it's not like you patented a better mousetrap, it's more like you patented the entire concept of rodent control. Now, if someone else builds a better mousetrap, you haul them to court.
Think about it, this is the issue with Amazon's notorious "one-click" patent. What they came up with wasn't particularly novel or clever, and because of their patent, the entire concept of "press a button to order something" is now locked away by one company.
Second, you have to bear in mind what the whole point of patents is in the first place--to promote the progress of science and useful arts. The deal is that if you're willing to write down your stuff on paper and make it public, then the power of government will protect your invention for a limited period of time so that you can make money off of it. This encourages you to do research and development and make stuff. Today, though, patent holders aren't upholding their end of the bargain. They're not using patents to promote progress of science and useful arts. No, they're using them to extort people into paying them large sums of money. They're using them to engage in extremely anticompetitive practices.
I mean, look at what is going on with the VP8 codec. MPEG-LA is essentially trying to claim patent on all streaming video technologies. It's not good enough that they have H.264, a decent codec. No, they want to make sure that no one ever comes up with a competing codec of decent quality. For years, they've been threatening that Theora infringes on their patents without offering up any proof. Now, they're threatening that VP8 infringes on their patents without offering up any proof. I can't find the quote right now, but I think I remember someone at MPEG saying at one point that they believe that any streaming video technology will likely infringe on at least some of MPEG's patents. This is most definitely not what the Founding Fathers had in mind when they set up patents to promote the progress of science and the useful arts. Indeed, it is significantly hindering those ideals.
I wish there were some way to meaningfully reform the system, but I just don't see anything on the horizon. In 2010, patents increased by 30%, and I'll bet a wad of money that a very large proportion of those were software patents.
Software patents are evil, plain and simple. They need to go. Most of them are patenting obvious things, many of them are patenting algorithms (which are not supposed to be patentable). It flies in the face of what the purpose of patents are and are significantly hindering the progress of technology and competition.
MPEG themselves (not MPEG LA) also issued press release: "MPEG envisages royalty-free MPEG video coding standard" http://bit.ly/hFMdGd
H.264 is a standard all right, the standard of an evil empire. A patent encumbered standard is no standard at all - not in the way the rest of the world uses the term. All the complaints about VP8 pale in comparison to this fundamental fact.