Linked by Thom Holwerda on Sun 31st Jan 2010 14:20 UTC, submitted by lemur2
Internet & Networking Despite the recent interest in adopting HTML5's video tag, there is still one major problem: there is no mandated standard video codec for the video tag. The two main contestants are the proprietary and patended h264, and the open and free Theora. In a comment on an LWN.net article about this problematic situation, LWN reader Trelane posted an email exchange he had with MPEG-LA, which should further cement Theora as the obvious choice.
Thread beginning with comment 407370
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[3]: costs
by lemur2 on Tue 2nd Feb 2010 09:09 UTC in reply to "RE[2]: costs"
lemur2
Member since:
2007-02-17

"There is the point, as has already been mentioned more than once on this thread, that Theora is based on the VP3 codec, and that Xiph.org have obtained an irrevocable royalty-free license for the VP3 patents so that they can develop and distribute Theora.

You read it right ... Theora itself is based on patented technology. On2 are the owners of that patent.


Well, On2 is the only known owner of Theora related patents. You can't patent a video codec as is and get a stamp of approval from USPTO that this is the one and only patent covering the codec.
"

Well, actually, as it turns out, that is exactly what the grant of a patent by the USPTO is supposed to be a stamp of approval for.

What you can do is to patent parts of codec and maybe some relations between the parts but you can't sort of patent it all. Besides, a video codec is very likely to include some well-known parts such as DCT and motion estimation which you even can't patent because of prior art or existing patents.

This makes it quite likely (not in this specific case but generally) that even if you had some patents on your product, it doesn't mean that it wouldn't violate somebody else's older patents. Besides even your patents could be invalid because USPTO generally does a lousy job and many patents are invalided later by a court.


No. Firstly, what is being patented is the method of compression of digital video data. The idea of "codec" (as in encoder/decoder) itself is quite old and therefore un-patentable. Heaps of prior art.

However, a few years before 2001, the methods of compression of digital video were all new. If there are earlier patents that made the same claims as to the method of compression as were made with On2's patent application for VP3, then the USPTO would not have granted the patent to On2.

Therefore, there were no parts of the compression methods of VP3 that On2 could not patent because of prior patents. These would have turned up in a patent search, and On2 would have had to be paying someone a license for those technologies. They weren't. The fact that they weren't paying anyone a license for earlier applicable patents is precisely what allowed On2 late in 2001 to give the VP3 patents to open source in the first place.

If there were previous applicable patents, and parts of VP3 were licensed from some other party, then VP3 wouldn't have been of any use to open source in the first place.

So there are no applicable prior patents.

As for the other parts of the method in VP3 that On2 could not patent because of prior art ... it follows that no-one else could have a prior patent for that either, also because of the selfsame prior art.

If anything, on the balance of probabilities, given the age of VP3, and the time-frame when digital video compression suddenly became something one would want to do, On2 are far more likely to be in the position of the ones to become a patent troll, rather than be the ones who are trolled against.

Remember the vintage ... On2 gave over the VP3 technology, including applicable patents, to open source in late 2001. That meant that in late 2001, VP3 was already seen by On2 as "obsolete technology". That in turn meant that the timeframe for On2 getting these patents in the first place is back in 1995-1998 period. Windows 95 did not even ship with a TCP/IP stack ... so much for the scope and presence of the Internet in that timeframe.

Edited 2010-02-02 09:20 UTC

Reply Parent Score: 2

RE[4]: costs
by _LH_ on Tue 2nd Feb 2010 10:50 in reply to "RE[3]: costs"
_LH_ Member since:
2005-07-20


However, a few years before 2001, the methods of compression of digital video were all new. If there are earlier patents that made the same claims as to the method of compression as were made with On2's patent application for VP3, then the USPTO would not have granted the patent to On2.


I strongly disagree. All modern video codecs trace their technology back to the eighties. The fact that people didn't download Divx movies before circa 2001 doesn't mean that the technology wasn't there.

The first somewhat popular video coding standard is H.261 from year 1990. It feature among others:

* blocks
* DCT
* quantized transform
* zig-zag scanning
* motion vectors
* entropy coding

All of these are included in Theora. I just checked from the Theora spec.

Nothing prevents other companies holding some obscure implementation patents on these technologies and even if On2 had done their best to check that they don't violate anybody's patents, that does not guarantee that there are no patents.

You seem to constantly imply that there exists an On2 patent which covers the VP3 as a whole. That doesn't hold. On2 has just guaranteed that if they have some patents related to VP3, then they won't use those against Theora.

Reply Parent Score: 2

RE[5]: costs
by lemur2 on Tue 2nd Feb 2010 11:06 in reply to "RE[4]: costs"
lemur2 Member since:
2007-02-17

"
However, a few years before 2001, the methods of compression of digital video were all new. If there are earlier patents that made the same claims as to the method of compression as were made with On2's patent application for VP3, then the USPTO would not have granted the patent to On2.


I strongly disagree. All modern video codecs trace their technology back to the eighties. The fact that people didn't download Divx movies before circa 2001 doesn't mean that the technology wasn't there.

The first somewhat popular video coding standard is H.261 from year 1990. It feature among others:

* blocks
* DCT
* quantized transform
* zig-zag scanning
* motion vectors
* entropy coding

All of these are included in Theora. I just checked from the Theora spec.

Nothing prevents other companies holding some obscure implementation patents on these technologies and even if On2 had done their best to check that they don't violate anybody's patents, that does not guarantee that there are no patents.

You seem to constantly imply that there exists an On2 patent which covers the VP3 as a whole. That doesn't hold. On2 has just guaranteed that if they have some patents related to VP3, then they won't use those against Theora.
"

AFAIK there are a group of patents held by On2 that cover VP3. These were all licensed to open source in late 2001.

http://en.wikipedia.org/wiki/On2
In late 2001, On2 released their VP3 compression technology into the open-source community including their patents on the technology. The technology lives on in the form of Theora.


The technology that became VP3 was developed some time prior to 1995.

AFAIK, at no time in On2's history have they paid anyone else for patented technology in VP3 belonging to someone else. Neither has any patent claim ever been raised by anyone against Theora.

It is exceedingly unlikely that there is a "submarine patent" out there, older than VP3, which can bring down Theora now.

Outlandishly unlikely. Astronomically unlikely.

If someone could have brought down Theora, they certainly would have tried it on by now, just as they tried (unsuccessfully) with Vorbis.

http://www.xiph.org/about/
The debate about whether or not Fraunhofer was within their rights or not is beside the point; this is an illustration of the amount of control commercial entities will attempt to exert over commodity standards


This is the whole reason why Apple et al are so much against Theora in the first place.

Edited 2010-02-02 11:16 UTC

Reply Parent Score: 2