That's a good point. Here is the payment history:

The maximum annual royalty (“cap”) for an Enterprise (commonly controlled Legal Entities) is $3.5 million per year 2005-2006, $4.25 million per year 2007-08, $5 million per year 2009-10, and $6.5 million per year in 2011-15.

…and you're right that there is a footnote saying. "Annual royalty caps are not subject to the 10% limitation" which is important.

I do find it a bit strange to hear people say the license fees are unknown. Even if you assume an unlimited cap, you can schedule out what the pro-rated fees will look like for you for the next 5-10 years based on your projected growth. You can argue that they are too high, but I don't think you can argue that they are unknown. If nothing else they are bounded on the high end.

Are you saying that last year, you could have predicted the licence fee cap would have been $6.5 million and not 5.5 or 7.5 or 10.5? If you couldn't then it was unknown until they announced it. Similarly, what do you predict it will be in 5 years and each of the next 20 years till the patents run out? Are you sure?

No, I'm agreeing that you're right that the cap is effectively unknown. I missed that footnote (shame on me). My point is that even with an unknown cap, the license rates are predictable for the foreseeable future:

    Free under 100k units
    $0.20 per unit above 100k
    $0.10 per unit above 5 million units

Now, I'm no fan of software patents and I wish they would be done away with. But given the choice between that and the practical/technical/patent uncertainty of webM I don't feel like those are massive payments.

I'm glad Google open sourced the on2 codec and I can even see that maybe the H.264 prices have been influenced by that move. But from a practical standpoint, I think Apple and Microsoft are making a reasonable move. People say it's short-sighted, but I think they are relatively protected.