Google called the MPEG-LA's bluff, and won | Hacker News Reader

Google called the MPEG-LA's bluff, and won

osnews.com

140 pointsthomholwerda88 comments

Saying that Google "won" is an absurd statement. What was the competition? What Google did was smart. The fear of patent infringement was likely the biggest issue VP8 had. They've taken that off the table.

But saying they've won seems to imply that someone lost. MPEG-LA didn't have to make a deal with them. But they got enough money that is was worth it to them. As far as they're concerned they get money from Google now, plus all the fees from everyone else. That's a "win" for MPEG-LA.

Both Apple and MS pay for H264. I'm sure both would be happy to not have to pay for a codec (although H264 will be around for a long time, even in the worst case). Microsoft's position on VP8 has been pretty clear... they're not shipping it (but maybe will now), but aren't opposed to it.

This was a smart move. But doesn't mean that there are winners and losers. And certainly there was no bluff that was called. Paying to license patents from a patent holder is hardly calling their bluff.

VP8 went from (according to the MPEG-LA and its proponents) legally dangerous to a better option (at least legally) than H.264, covering more users and more usage scenarios (namely, not just non-commercial, but commercial too), and for free to users, to boot.

So yes, VP8 (and thus Google) won.

That's like saying someone claimed they didn't order something at a restaurant while the restaurant claimed they did. The person pays anyway and the restaurant is happy and they don't get in trouble. Did the person win? Google has claimed for a long time VP8 was clear of patents. This is them giving up.

Doesn't it depend on how much they paid? If the person paid for much less than the restaurant was asking and/or for less than a court case would cost, I'd say they won. And as long as the size of Google's MPEG-LA payment isn't visible in their next earnings report (remember the MPEG-LA was claiming hundreds of millions of willful infringements), I'd say Google won, too.

> Doesn't it depend on how much they paid? If the person paid for much less than the restaurant was asking and/or for less than a court case would cost, I'd say they won.

I guess that's a matter of opinion then.

Because if the person really didn't order (or receive) any meal, and they still had to pay anything to clear up matters, I wouldn't say they won. At most I'd say they got off relatively well, considering circumstances. The restaurant, however, getting any money for a meal they didn't make, is a clear winner.

On the other hand (and I'm not sure how/if this holds up in the analogy because I'm not entirely clear about what's going on with the Google/MPEG-LA thing) if the restaurant did prepare the meal they claim the person ordered (but who didn't), assuming they got less than the costs of the meal, both are losers. Which figures because miscommunication often lead to waste.

The analogy breaks down when the agreement applies not only to Google, but also everyone else using VP8/9.

So yeah, Google may have paid the restaurant for their meal but now everyone else who goes to that restaurant and orders the same thing now gets it for free.

If you're hungry, I'd recommend ordering what Google ordered and eating for free.

As much as I wish this were the case and admire VP8 as you do, I don't think this is really true yet:

>covering more users and more usage scenarios

given that I'm not aware of any phone, set top box, AV accessory, etc that supports VP8 accelerated decoding.

From the list http://wiki.webmproject.org/hardware/arm-socs I would say some devices on the market should already support hw vp8 decoding.

Hm. I'd love to be wrong, I'll try some of them when I get home with my webm videos.

From http://wiki.webmproject.org/hardware/devices it looks like that the raspberry pi support hw decoding (edit: but maybe it's not supported at the driver level...)

Ah, looks like the only thing I'd have off that list is the Samsung Nexus 10 and it's packed away in a box. I presume it being on that list means it works, still a pretty tiny fraction. I'm sure that will grow, especially after this good news.

Nobody has lost because the fight has moved on.

VP8 being a safer choice isn't reassuring, since it isn't much of a choice anymore.

Welcome the new fight, VP9 vs. H.265 vs. Daala.

This agreement covers VP9 as well.

Only to the extent that it uses the same techniques as VP8. Anything new (read: anything which might infringe on H.265 related patents) is still subject to threats from MPEG-LA.

Yes, but this applies to any and all software - including H.26{5,6,7,8,9}.

Sadly.

In his response to this MPEG-LA news Xiphmont (in response to a troll comment) claims Xiph are contributing to VP9, including writing the spec, so it may well be that VP9 aka VP.next is merging with Daala and the IETF Internet Video Codec project. It makes sense to me as Google and the IETF both provide some level of defence against all this patent BS, which would be pragmatic for Xiph to take advantage of.

Strongly disagree.

> As far as they're concerned they get money from Google now, plus all the fees from everyone else.

They get money from Google now (presumably, but we don't know the details), and they get money from h264 users- but they do not get anything from any other VP8/9 users. H264 users are now paying for something that VP8 users get for free.

Just when h265 is nearing release, MPEG-LA is effectively giving up all claims to the competing technology- VP9. I see no way that this does not eat into h265 adoption and h265 is going to have to be technologically far superior to VP9 to justify the licensing fees.

> doesn't mean that there are winners and losers

There's a clear winner- you and I. There's now a completely free alternative to h264.

> completely free alternative to h264

As a user H.264 was already free. As a developer H.264 was already free.

There are licensing costs built into OSes shipping h264, devices built with HW decoders, online movies rented with it, etc.

So by having VP8 I save a few cents off the price of an OS upgrade or new device.

And in exchange I get significantly poorer quality video.

For every copy of Windows 8 Basic sold, Microsoft saves $2 by not including. That adds up!

http://www.extremetech.com/computing/142796-how-to-play-dvds...

Nothing is stopping you from paying for h.264 for better quality (although the difference is far from what you try to portray).

Meanwhile those who doesn't want to pay for the priviledge of the better quality h.264 offers (and likely h.265 in comparison with vp9) will be able to use a free codec, hopefully one which will be standarised under HTML5 and thus supported across all HTML5 compliant software.

See, it works out for everyone.

And if you're a content creator or video professional?

Or try to make a website that serves a lot of H.264 video and then come back and tell us how free H.264 is.

More generally, this is a video format we're talking about. "Free as in speech" is a huge deal. Speech is what it's for, for crying out loud.

Depends on what you are using it for e.g. renting TV shows or for cut scenes in video games.

They also reserved the right to start charging for free videos (you remember when Apple fought against the MPEG-LA when they tried that with AAC audio codec, right?). They only made that permanently free a few months after WebM was announced.

They've taken that off the table.

If only patents worked that way.

Read the linked article: http://xiphmont.livejournal.com/59893.html to see how exactly Google has won.

How does agreeing with a random third-party about how you can use your own product, and worse, paying them for the mere unproven possibility that they own parts of your product, count as "won"?

In any other situation, history would classify this as a clear cut case of racketeering or extortion.

Yes I am still waiting for that answer. Maybe I am being dense but to me this is not a victory? If anything seems like a loss? In other words suspicions now are founded that VP8 was never free of patent infringement.

It's arguably a victory within the existing rules of the game. Taking it to an extreme - if Google paid a single cent in order to remove the threat of MPEG-LA related patent suits and in return massively increased adoption of VP8 and following specs, would that have been a victory? A dollar? Ten dollars? A million dollars? It's a victory for Google if the cost to them is less than the expected benefits of removing a barrier to adoption.

Of course, the flipside to this is that it's also a victory for the existing patent system. It's usually cheaper for people to settle than it is for them to fight a patent suit, even if they win. At the end you maybe get most of your costs back, and it's cost you several years of your competitors telling your potential customers that you're a patent infringer and if they implement the spec they could be next.

So yes, a real victory would require certainty over the patent status of VP8 without Google having had to pay anyone anything, but also without having to spend years in litigation to prove that. But that's not an option at the moment, and the fact that Google were able to negotiate a relatively wide-ranging and liberal license (without any admission of infringement) is arguably the best that's possible without significant changes in the law.

haha this is absurd, and drawing the conclusion in the headline makes about as much sense as "Google admits its WebM (VP8) codec infringes MPEG H.264 patents; agrees to license technology" [1]

It is so hard to take OSNews seriously.

[1] http://macdailynews.com/2013/03/08/google-admits-its-webm-vp...

I think the article you linked is a pretty reasonable interpretation. Google was claiming VP8 didn’t infringe any patents they didn’t own, which apparently wasn’t the case, and now they’re having to backpedal and say, “well, it did infringe them, but we’re covering the licensing.” My understanding is that MPEG-LA does not engage in “patent trolling,” but exists to provide an easy way to license all the (known) patents. I think it would be up to the individual patent holders to try to sue for infringement. I think Daniel Eran Dilger is spot-on here; there’s been a lot of misinformation being spread by some proponents of free video codecs, first that Theora is good or almost as good as H.264 based on some crappy codec comparisons (it’s not), and then that VP8 was not patent-encumbered. I recall thinking the situation with VP8 when Google first released it for free kind of seemed to me like a “turd sandwich,” since the quality was somewhere between slightly inferior and very inferior (due to a bad encoder) to H.264, and the licensing was uncertain, despite Google’s claims.

It seems that google's position is firmly that this agreement is not an admission that they infringed any patents. That said, it's a pretty funny statement coming from an entity that just paid for patent licensing. The tricky thing about the situation is that the details of the agreement aren't clear enough to say who "won", and personally, it looks to me more like a stalemate. Even stalemate might be putting it strongly, as it implies that they are in some kind of opposition, which may not be the case.

"Paid for" is an assumption isn't it?

Oh you're right. It seems they are being tight lipped about whether money was exchanged.

Google was claiming VP8 didn’t infringe any patents they didn’t own, which apparently wasn’t the case, and now they’re having to backpedal and say, “well, it did infringe them, but we’re covering the licensing.”

VP8 saw little market acceptance because of the legal threats of MPEG-LA (did that consortium ever actual detail specific "infringed" patents? The last I saw they just made blanket statements that there must be something). So Google paid them protection money. By no analysis does this mean that they admit infringing of anything, and such an analysis of legal agreements is woefully naive.

I don't understand why so many comments are painting this as bad. Who cares if Google paid a shitload of money to MPEG-LA. The main point is, VP8 is now free for anyone to use, for any reason. Wasn't that the goal?

It's a real technical "win", but it doesn't look that way. It's just a marketing problem now. Google needs to spend their resources pushing VP8 into everything now.

No, VP8 is not necessarily "free", there are numerous companies that may wish to sue the users of VP8 over patent infringement. Qualcomm, Broadcom, Samsung etc. (Why would you want a free alternative when you're the only few companies that can make hardware encoders?)

Companies that are in MPEG-LA won't sue, but there are many "relevant" companies that have not joined. Companies are still being sued (and settling) over patents supposedly infringed by H.264.

The system is a mess. Google are moving in the right direction by eliminating a large potential threat (MPEG-LA), but it's job is not done yet.

Note that Google pushing VP8/9 is trying to promote competition and reduce the licencing costs of H.264/5. This is very similar to what Microsoft (rightly) did before with VC-1.

Samsung couldn't sue - they've already shipped hundreds of millions of WebM implementations that they lose their patent license for if they do. Depending on how much WebM-related code ended up in their graphics drivers and/or how widely distributed their Android reference platforms are, Qualcomm could easily face similar issues. And well, um, Google has also clearly taken care of a third major H.264 patent holder outside the MPEG-LA - Motorola. I'm sure there are others out there, but the pickings are getting thinner and thinner.

> Samsung couldn't sue - they've already shipped hundreds of millions of WebM implementations that they lose their patent license for if they do.

I don't understand, why would they stand to lose their patent license for WebM?

It says on Wikipedia about WebM that "The project releases WebM related software under a BSD license and all users are granted a worldwide, non-exclusive, no-charge, royalty-free patent license" so how can anyone lose the right to that license?

And given that the above WP quote is in fact true, what would Samsung sue over in the first place?

> And well, um, Google has also clearly taken care of a third major H.264 patent holder outside the MPEG-LA - Motorola.

Because, what did they do to Motorola then?

Would I be sort of correct in saying this whole matter is more about power/politics between these big corporations (using patents as leverage) than it is about the actual legality of who is doing what with whose algorithms--in the sense that they do use these legal tactics to attain their goals but the usual goal (justice) is secondary?

That'd explain why it's so confusing to me, laws are usually quite logical (especially the codified type), but politics and power gets really complex really fast (to me).

> why would they stand to lose their patent license for WebM?

From http://www.webmproject.org/license/additional/ and http://www.webmproject.org/license/bitstream/

> If you or your agent or exclusive licensee institute or order or agree to the institution of patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that this implementation of VP8 or any code incorporated within this implementation of VP8 constitutes direct or contributory patent infringement, or inducement of patent infringement, then any patent rights granted to you under this License for this implementation of VP8 shall terminate as of the date such litigation is filed.

Right. That explains that then, thx :)

And somehow all these corporations (Samsung, Broadcom, etc) believe they have a claim to these patents? But it can at best be only one of them, right?

>there are numerous companies that may wish to sue the users of VP8 over patent infringement. Qualcomm, Broadcom, Samsung etc. (Why would you want a free alternative when you're the only few companies that can make hardware encoders?)

What are you talking about? Nobody is going to use a software encoder where they need a hardware encoder. The whole point of a hardware encoder is that doing it in software won't fit within your processor consumption or energy budget. And hardware manufacturers will be happy to sell VP8 hardware encoders to anyone who wants to buy one -- especially now that they know they won't have to pay MPEG-LA any royalties to do it.

On top of that, Qualcomm and Broadcom are WebM supporters.

http://www.webmproject.org/about/supporters/

TL;DR Google gave MPEG-LA a bunch of money. Yep, Google sure showed them!

If Google have MPEG-LA $1million just to go away, then yes they probably won. A single patent case defence would cost them at least that much, plus they get to inoculate VP8 and VP9 against MPEG-LA patent pool FUD PR.

Without knowing the amounts involved, it's impossible to really say who "won". Google obviously got what they wanted: indemnity for the MPEG-LA patent pool for VP8 and future codecs based on it. Maybe that cost them $1billion, maybe it cost them $1million: we don't know.

Maybe I came from a different era, but the party that pays is generally considered the loser. Strategically, Google and MPEG-LA one. Google got the indemnification they swear they didn't need, and MPEG-LA got money and further proof that their patents were that strong that not even the mighty Google could ignore it.

This is the same strategy patent troll, rent seekers use. It doesn't matter if they get $1 or $1mm from Microsoft or Apple. To everyone outside the negotiations they are clearly the winner no matter how flimsy the evidence looks.

> Maybe I came from a different era, but the party that pays is generally considered the loser

Or maybe you're talking absolute nonsense as you clearly don't understand what the agreement is about.

You misunderstand the nature of the agreement.

Please point to a case where MPEG-LA has given a completely sublicensable royalty-free right to someone for H.264, for any amount of money.

Can you get all the H.264 rights MPEG-LA offers from microsoft or apple, without going through MPEG-LA? Does it cover all use cases?

H.264 is obsolete. MPEG-LA is moving on to H.265, the new hotness. It is telling that MPEG-LA's sub-licensing for VP8 only covers one subsequent revision of the codec. If VPx+ is augmented with H.265'ish technology in order to stay competitive, we'll see a new round of fighting.

You completely ignored the question I asked, by raising an irrelevant point :)

I'll point out that despite your claim that H.264 is "obsolete", nobody still has gotten the kind of licensing google just did, despite people trying.

Both Microsoft and Apple have patents in the H.264 patent pool. It is in their interest for H.264 licensees to go through MPEG-LA; it is not in their interest to give away sublicenses.

Actually, neither make much revenue from that pool. But, fine, pick someone else. AFAIK, nobody has the right to sublicense all the H.264 patent rights to others.

This is one reason, for example, you have cameras and camcorders that ban professional use.

To be fair, Google did not get this either; this deal does not give them the power to sublicense H.264 royalty-free.

It gives Google the power to license VP8 to others without fear of patent suits from the MPEG-LA. But VP8 was already Google's codec to license. And they have spent years claiming it was not patent encumbered.

So basically it looks to me like Google is paying to do something they always claimed they could do anyway.

Yes, you are essentially correct. However, my point was that nobody has the right to do this in the H.264 world, so it is in fact, still significantly different.

Meanwhile, in reality, everyone is using H.264. Seems like a lot of wasted effort to me.

Well, next up is Google paying off the Scene to drop H.264 and switch to VP8 :)

If you cut the crap, does this mean VP8 is safe to use for everyone for free (without royalties to either MPEG-LA or Google) or not?

Safe except for not working in most places you care about.

I know many here don't like/trust Florian Mueller but he is claiming that Nokia (not part of MPEG-LA patent pools for either VP8 or AVC) is currently suing HTC for patent infringement regarding VP8 in Germany. Google is acting as an intervenor. Assuming this is correct (and I have no reason to doubt it) VP8 is not in the clear just yet and Nokia did not participate in the standardisation so is not committed to FRAND licensing.

It is obviously possible that the court could rule the patent is either invalid or not infringed but I think it would be wise to avoid counting chickens just yet (unless anyone has good alternative and contradictory sources on this).

http://www.fosspatents.com/2013/03/patent-clouds-remain-over...

It's a shame. As much as i wish this headline would be true, it does only seem to me that Google bent to MPEG-LAs will, a conglomerate of Apple, MS and what not..

From my point of view, the MPEG-LA needs to be shut down, not subsidized with more money. It shows every sign of a typical patent troll and will always choose the option that guarantees its members income, not the most innovative or best option, but the one where there the most money can be leeched out of people and companies.

"Google comes to comprimise agreement with MPEG-LA which satisfies no-one's tribal blood lust."

It's like negotiating with a terrorist.

Shutting down the MPEG-LA is an absurd idea. Instead of one point of contact every licensee would have to deal with each patent holder individually raising the cost further and increasing the chance of litigation down the road. The reality is it has helped Google and many other companies adopt MPEG tech.

Not an absurd idea. The problem is the company and money driven structure not the MPEG-LA per se. Replace it with a consortium that has a general interest in innovation and not money.

Who would be in that consortium. Companies that have a "general interest in innovation and not money?"

Sorry to tell you, but that set of companies is: {}.

If MPEG-LA was run the way it was actually required to be run by the DOJ (http://www.justice.gov/atr/public/busreview/215742.htm), we'd be better off.

They are supposed to be getting independent experts, etc.

The current "independent expert" has the following qualifications: "The expert helped form the MPEG-LA, helped in drafting the first MPEG-LA licensing agreements, answers questions from licensees on behalf of the MPEG-LA, has attended business settlement meetings on behalf of the MPEG-LA, and has testified before US congress on behalf of the MPEG-LA. Heck, he is listed on the MPEG-LA website as "MPEG-LA's US patent counsel"."

Your set should at least look like this: {Mozilla}.

On the one hand, MPEG-LA appears to have "won", because they got Google to licence their patents. On the other hand, the license Google has acquired is incredibly broad: MPEG-LA usually seems to offer broad non-commercial / small income rights but reserves the right to charge considerably more for large scale commercial use of codecs that they regard as infringing patents in their pool. They've given up on receiving any of that income from future use of VP8.

Perhaps they regard VP8 as a dead codec, now that h.265 is on the horizon, so not worth the licensing effort?

The agreement covers VP9 too.

The license agreement also applies to future codecs derived from VP8, most specifically VP9, so I doubt the MPEG-LA thought it was giving a license to a dead effort.

Point. I think the general thought still stands: Perhaps MPEG-LA believes they have now established h.26[45] strongly enough that VP9 will never be a threat.

Of course, if we knew how much Google had actually paid for these patent licenses we'd have a much better idea of who, if anyone, had actually "won"! I doubt we'll get that information though.

The alleged infringement was massive - every download of Chrome 6+ and every Gingerbread+ Android device (or at least the ones with the Google apps) to start. And all of it could easily have been willful - On2 and MPEG-LA had apparently tangled long before Google got involved.

Given that, I think not finding out financial terms is a tell all by itself. As I've said elsewhere in this thread, I don't think Google would be able to hide the cost of an MPEG-LA win in their earnings reports.

Not to mention that the free internet broadcast license of h264 expires in 2015. After that I am fairly sure that Google will be looking at hefty fees for YouTube if they do not convert over to VP8.

The MPEG-LA extended that free licence to forever already due to the threat of vp8. And YouTube is already converted to VP8.

On the other hand you still need to pay to sell TV shows and movies, which Google has just started rolling out as VP8 (starting with ChromeOS clients but expanding from there).

Google Play Movies and TV is going VP8? Interesting.

> It shows every sign of a typical patent troll and will always choose the option that guarantees its members income, not the most innovative or best option,

1) It's members are the ones that invented h.264 (NTT, Microsoft).

2) Every company that owns something of value tries to maximize their revenue from it.

MPEG-LA is a perfectly reasonable arrangement. It allows things like h.264 to be standardized while also offering an avenue for that R&D work to be monetized. The alternative would be people keeping new coding techniques secret and trying to compete with each other using them to offer better video quality in their own products. There is not a world where Microsoft, NTT, etc, are going to do R&D to help competitors like Google out of the goodness of their hearts, for the sake of "innovation."

It should be noted here that Google makes a ton of money off video compression algorithms (they make Youtube possible in the first place), but has done jack shit to contribute to their development itself. Of course they want to get the license to something like h.264 for free, but why shouldn't they have to pay for it?

They did pay. Google spent over $100 million on On2 Technologies, the company behind the VP codecs. Did you expect more than buying a company and employing their developers?

What does that have to do with H.264?

GP says Google wants something "like" h264 without paying for it. Somehow paying $100M for the company that developed the codec, and employing its people to continue development for years now, doesn't count. rayiner: I think you made some good points but the unfortunate rant at the end, that's just anti-Google bias...

Regardless of whether Google "won" or not, it does look like VP8 will be a much more approachable codec thanks to being open source, and because it doesn't require video makers to pay MPEG-LA license fees whenever they want to make money from their videos. I'd think at the very least professional camera makers would be interested in supporting it. From that point of view, I think the article is right.

The real win will come if Google is able to have VP8 adopted as an MPEG standard, like the article states.

>it does look like VP8 will be a much more approachable codec thanks to being open source

How is VP8 'open source'? It is a standard. There exist open source implementation(s). However there are open source implementations of H.264 decoders and encoders as well, it even has a reference implementation [1] that anybody can use[2] (patent issues aside, but (understandably) very few open source projects offer the users patent indemnification).

Is it any wonder that most non-tech people are confused about these issues when even people in the tech community don't seem to understand the differences between standards and implementations, between 'open source' and 'free as in beer' and between secret proprietary formats and open standards? Sadly we seem to be at a point where 'open', 'standard' and 'free' are just words thrown about to mean 'something I like'.

[1] http://iphome.hhi.de/suehring/tml/download/ [2] http://iphome.hhi.de/suehring/tml/download/COPYRIGHT_ISO_IEC...

thanks to being open source

Something can be open source and still have patent threats.

That's one of the big differences between patents and copyright. If I go into a room and don't talk to anyone and independently write some software, then no-one can sue me for copyright infringement, but they might be able to sue me for patent infringment.

Even better, given that the issues raised by (iirc) Mircosoft and Apple against webp becoming the HTML5 standard video codec (and thus mandatory to support for HTML5 compliance) was that of 'patent uncertainty', with this agreement we can finally get a really great royalty free standard codec which can be used with any browser which is HTML5 compliant.

Same goes for WebRTC, which should now be on the fasttrack to be accepted as the standard for RTC (currently supported by Chrome, Mozilla and Opera).

I think having open and royalty free standards for such integral parts of what is likely to be the 'future of the web' is extremely important to foster innovation, thanks to VP8/VP9 and Opus we now have technically competitive offerings which anyone can implement free of charge across all platforms and for whatever interesting solutions they can come up with.

> the issues raised by (iirc) Mircosoft and Apple against webp becoming the HTML5 standard video codec

"webm" is the video format (and uses VP8). "webp" is a related image format.

[webp is not very compelling as an image format on its own, but maybe the relationship with webm confers some advantage...]

yes, that was a spelling error, but given the context of the discussion (standard video codec) I doubt anyone failed to understand that I meant webm.

Will google now follow through and remove H.264 support from chrome?(1)

(1) http://arstechnica.com/information-technology/2011/01/google...

Reply on news.ycombinator.com