Google Opens Up VP8, Launches New Container Format

[deanjo]

From the looks of it (the crappy state of the software), Google didn't work on the codec. They probably needed all this time to verify that VP8 doesn't infringe on MPEG patents.That's an easy job, but takes time.

You think they would have done that with Google Talk and Google Book search as well don't you? Didn't work out so nice and clean there now did it?

Compare just-opened VP3 with Theora today. It's miles apart.

Sure they are, and how long did that take to get to that point?

[smitty3268]

I never said that, I said it was an important factor. I did not say the most.

I said quite the opposite. I mentioned MS and Yahoo. Those are hardly a "geek in the basement".

I apologize if I came off a bit snarky there, that wasn't my intent. Too much sarcasm doesn't translate well into text.

I was laughing awfully hard at some of those posts you wrote, though.

Ya, just like they just caught their wifi data collection that happened years ago.

Correct me if I'm wrong, but did Google just spend 6 months saying they didn't do that, only for us to then find out that they were wrong and actually did? It was my understanding that this was something that just happened while nobody noticed, and as such is a completely different scenario.

Lawyers by the way don't make the decision on patent infringement. They present their argument. The rest is up to judge or jury, usually filled with unqualified individuals who have a hard time grasping the technical aspects.

Sure, but patent infringement claims generally aren't that grey. You either infringe or you don't, the real question is whether the patent is even valid in the first place or should be invalidated. And I think you might be surprised by how good a job juries typically do. There's always the famous screwups, but the majority of the time they get things right.

I wouldn't be shocked at all. With the amount of patent trolls out there and google putting it out as opensource it makes it all that much easier for them to pick it apart. It's not only the mpeg-la that one has to worry about. What could be juicer then going after one of the largest companies in the world with deep pockets who laid out everything in plain sight?

Please read my comment again. That's what I said - that some unknown patent troll could come along with something. What's unlikely is for it to come from an existing mpeg-la patent that is already well known and therefore specifically worked around by VP8.

[deanjo]

Correct me if I'm wrong, but did Google just spend 6 months saying they didn't do that, only for us to then find out that they were wrong and actually did? It was my understanding that this was something that just happened while nobody noticed, and as such is a completely different scenario.

It came after a German data protection authority requested an audit. Had no audit been requested the issue would have never seen the light of day.

Sure, but patent infringement claims generally aren't that grey. You either infringe or you don't, the real question is whether the patent is even valid in the first place or should be invalidated. And I think you might be surprised by how good a job juries typically do. There's always the famous screwups, but the majority of the time they get things right.

There are literally thousands of patents out there that are vague. That is a shade of gray. Patent trolls love these patents because they are subject to a lot of interpretation.

Please read my comment again. That's what I said - that some unknown patent troll could come along with something. What's unlikely is for it to come from an existing mpeg-la patent that is already well known and therefore specifically worked around by VP8.

You assume of course that all of MPEG-LA's (or their members) patents are known. Wouldn't shock me too see MPEG-LA launch a submarine attack.

[deanjo]

All I'm just saying is that adoption won't be overnight, far from it. With all the players involved on both sides there will do doubt be some legal issues. These tend to slow progress down and in this case it would only slow down the VP8 project but not slow down h264 and kin. To quote Dan Cook "The opera ain't over till the fat lady sings." and the fat lady hasn't even taken the stage yet.

[smitty3268]

It came after a German data protection authority requested an audit. Had no audit been requested the issue would have never seen the light of day.

So in other words, it's completely different. There has already been an audit here by Google, and nothing was found. So what does that have to do with another situation where nothing was audited at all?

Nothing, you're just trying to associate one bad experience to the company as a whole saying bad things will continue to happen. It's no different than if I pointed out past Apple failures (like the Newton) while talking about how the iPad will fail. The 2 are completely unrelated.

You assume of course that all of MPEG-LA's (or their members) patents are known. Wouldn't shock me too see MPEG-LA launch a submarine attack.

I was assuming that those members would have a strong motivation to announce all such patents they had, rather than keeping some around for submarine attacks, because they would receive a higher proportion of h264 licensing money if they had more patents as part of the pool. I could be wrong, though, maybe that's not how they decide to split the money.

[smitty3268]

It's also worth noting that if a troublesome patent shows up from a small enough patent troll, Google could just gobble them up.

Also, the MPEG-LA can't sue directly, they are just a common front for a variety of companies behind the scenes. While I'm sure a lot of the companies involved would indeed like to sue, it would be that individual company making a choice to sue and not some big conglomeration. There may well be many companies with patents in the MPEG-LA that are fine with Google's VP8 and would choose not to sue even if they could.

[smitty3268]

There may well be many companies with patents in the MPEG-LA that are fine with Google's VP8 and would choose not to sue even if they could.

Since that's not the conventional wisdom, let me explain a little more what i mean.

Let's take Microsoft as an example. As a member of the MPEG-LA, they make money for every license sold. However, they pay out far more than they make to other members, because they are a heavy licensee themselves in addition to being a licensor. In their particular case, they may figure the more licenses we can sell to others the better off we'll be, so they would have an incentive to sue.

However, another company may face a different set of calculations. They may feel like if they could replace their usage of h264 in their devices (lets say they make camcorders) with a freely licensed VP8 then they'd actually save tons of money. So it might make sense for them to just sit on any patents they have, rather than losing more money by paying for their h264 use.

I guess my point is that we shouldn't view the MPEG-LA as one monolithic entity when it comes to enforcing patent claims on others. Each member of the group will have to individually decide whether they want to fight a big, heavily publicized court case or not. The answer may well be yes, but then it could also be no. Especially if only 1 or 2 patents are found, google might even be able to pay hush money or freely license that patent to everyone else for enough money.

[deanjo]

It's also worth noting that if a troublesome patent shows up from a small enough patent troll, Google could just gobble them up.

Google hasn't been able to do that so far on previous IP suites.

Also, the MPEG-LA can't sue directly, they are just a common front for a variety of companies behind the scenes. While I'm sure a lot of the companies involved would indeed like to sue, it would be that individual company making a choice to sue and not some big conglomeration. There may well be many companies with patents in the MPEG-LA that are fine with Google's VP8 and would choose not to sue even if they could.

Actually you are wrong there. As a entity it can sue. They did just that agains Alcatel.

http://www.mpegla.com/main/pages/media.aspx

[Yfrwlf]

Works fine here with chromium. But it's a bit cumbersome. First you have to go here to enter the HTML5 beta:

http://www.youtube.com/html5

And then you have to append &webm=1 to your search queries.

e.g. http://www.youtube.com/results?search_query=foo&webm=1

IMO quality is just as good/bad as flash. The main thing missing from the player is an option to go full screen.

That would be because you're converting a h264 (lossy) into a V8 (lossy) = more lossy?

Correctly comparing codecs requires, at the very least, compressing both from the same source, and to be more fair the source needs to be very raw.

Definitely interested in some comparison tests.

[deanjo]

So in other words, it's completely different. There has already been an audit here by Google, and nothing was found. So what does that have to do with another situation where nothing was audited at all?

You want more examples look at Google Talk and Google Books.

Nothing, you're just trying to associate one bad experience to the company as a whole saying bad things will continue to happen. It's no different than if I pointed out past Apple failures (like the Newton) while talking about how the iPad will fail. The 2 are completely unrelated.

Not at all just pointing out that Google is far from infallible and thinking that they have done due diligence in researching is a big leap of faith.

I was assuming that those members would have a strong motivation to announce all such patents they had, rather than keeping some around for submarine attacks, because they would receive a higher proportion of h264 licensing money if they had more patents as part of the pool. I could be wrong, though, maybe that's not how they decide to split the money.

From what I understand of the MPEG-LA there is no member that has more pull then the other and are all equal partners.