I think Chris really gets it. Good blog post, enlightening on some levels.

In the spirit of being brief, I’ll preface my comments with a big “me too”. Personally, I think you hit the nail on the head — especially with the “bringing the web to video” vs “bringing video to the web” comparison. I think the same logic points to the problems with the membership of the consortium/working group/cabal/whatever in charge of creating the HTML5 spec. — their motivations.

Projects that begin in an open manner often seem to grapple first with the questions around “how can we make this possible” and successful ones are able to focus on this purely from the standpoint of how to deliver functionality and features to consumers (users). The focus is on developing a viable technology.

All you list have commercial motivations involved. It seems to be the internet equivalent of the “3rd-generation management curse” that open standards get perverted from “how can we make this possible” to “how can I create a protected revenue stream from this thing that people have glommed on to.” The focus is on creating revenue.

This is an inevitable (and desired) behavior in capitalistic economies, but if the focus is TOO MUCH on protecting the assets/content/revenue stream of the providers, and TOO LITTLE on the user experience of those meant to consume it, then it will fail, for both groups.

See I actually know I am not a Lawyer ;-P

No one can guarantee codecs (or anything else) are free of intellectual property issues. To do so you would need to be sued by every IP holder and win in court since the courts are the only one who can decide that. That is one of the things that sucks about the patent system. You can have a patent lawyer check to see if your code tramples on any known patents but then there are always submarine patents and also patents which are totally unrelated but someone sues you anyway (and costs you money). This is not unique to any class of codec (remember the .gif debacle?).

Xiph.org, the maintainers of of Theora have stated in their HTML5 position paper (http://www.xiph.org/press/2007/w3c/) that “Xiph knows of no infringing technology in Ogg. Tens of millions of copies have been deployed worldwide over the past ten years in a diverse array of software and hardware products from small .orgs like Wikimedia to giant commercial vendors such as Samsung and Microsoft. Ogg has triggered no litigation to date.”

On the other hand MPEG-4 which requires a per seat license including another per seat license for the audio codec used has had claims leveled against it. I encourage you to go read the position paper as it is written much better that I could explain it.

Also, the world does not need me video blogging. That would be a career mistake. :)

You have been doing a great job in your blogs and it would be great to hear even more about Mozilla’s work in supporting Theora via the tag. Also perhaps getting a Mozilla engineer to whip up some examples of how to distribute Theora files for FF3 while also falling back to other methods (plugins, other formats, etc) for legacy browsers. Perhaps you could video blog this stuff with Ogg ;-)

Thanks for the post. I always find them enlightening.

True the old encoder sucked but Monty of Vorbis fame has been working on Theora 1.0. In his honest evaluation of the Theora codec which he posts here http://web.mit.edu/xiphmont/Public/theora/demo.html, Monty states …the current encoder is lacking compared to the very very best. It certainly is. Yet, the Theora format is entirely capable of closing with H.264 and MPEG-4 in terms of R-D while still requiring a fraction of the CPU time.

In other words there is more work to be done but the Theora codec can be just as good in quality as H.264 and MPEG-4. In fact current code Monty is working with improves the quality quite a bit as most peoples perceptions come from the old first drops of code. As with anything in the world things take time to hit their stride. Theora is one of those and when Direc is available (estimates are about two more years) we will have two codecs which are open and redistributable unlike codecs based off of H.264 and MPEG-4.

Yes, it’s well known that Theora doesn’t have the same quality as some of the newer codecs (including the ones in HD!) As a friend of mine just pointed out, things like Dirac, which are under development, are targeted at HD-quality use cases.

But there’s a lot of room in the market for various kinds of quality. Webcams aren’t going to produce HD quality video any time soon, nor do I think that the users of those webcams want everyone to see every pimple in their faces. :)

And yes, the standards are made to last. It would be silly to make a web video standard that lasted only for a year or two. Couple years from now everything will be HD. It just is not possible with Theora.