On 2 Apr 2008, at 16:55, Robert J Crisler wrote:
It will be very, very difficult to develop critical mass for content encoded in Theora (or Dirac), much less ubiquity. I'm not saying there's no point in trying. I applaud the effort, though I have misgivings about the W3C setting itself up as a video/audio standards organization when we already have the Motion Picture Experts Group.
I don't think anyone whatsoever is suggestion to create a new codec — we'd gain nothing by doing so.
But ... why not recommend that web developers encode in MPEG-4 AVC or Theora?
MPEG-4 has patent fees to be paid, making it impossible for Firefox or Konqueror (for example) to comply to that.
Theora has unknown patent status, and big companies are unwilling to implement it (as it has little pre-existing content, and it is no better than what they already have) lest they get sued due to some submarine patent.
At least that would give some direction out of the current morass. ISO/IEC standards, like AVC/h.264, are vastly preferable to single- vendor (non)standards from Adobe, MS and Real.
All the codecs that have publicly been looked at already have glaring issues with actually getting them interoperably used. We need something everyone is willing to implement. If people don't implement what we say, what we say is irrelevant.
Why should the W3C choose not create a better situation than the current one (which is a mess for developers and a mess for users), while continuing to work on the ideal?
There's a reason why the status quo is the status quo: different people willing to implement different things. One standard cannot force people to implement something they don't want to. We cannot just create a better situation: people have to actually do what we say to be in any better situation than we already are. One group can't implement specifications with known patents, and the other is unwilling to implement specifications with no known patents, due to submarine patent risks.
-- Geoffrey Sneddon <http://gsnedders.com/>
