RAND, patents and VoiceXML 2.0

All -

In my professional capacity at the MITRE Corporation, I lead a team 
working on infrastructure for advanced research in spoken dialogue 
systems. As a result, I have had the opportunity to review the 
VoiceXML 2.0 proposal as it was developed, since my team has a 
representative on the W3C Voice Browsers committee. I expect that we 
will address our technical concerns as a group; but that's not why 
I'm writing. I mention this only because some small subset of readers 
might recognize my name, and I want to make it clear that I'm not 
writing in a professional capacity. What follows is my personal 
opinion.

Whatever the technical value of the VoiceXML 2.0 specification might 
be, the patent claims made by the committee members render the 
specification unfit to be a W3C standard. I am aware that the W3C is 
debating this issue even as we speak, and it's possible that the 
VoiceXML 2.0 specification may be one of the first test cases for 
whatever policy the W3C adopts. Even the possibility of granting 
standard status to proposals for which "reasonable, 
nondiscriminatory" fees may be charged due to patent claims would be 
anathema without the US Patent Office's lunatic softheadedness about 
software patents. If the W3C, which is supposed to be a world-wide 
consortium, surrenders to this sort of insanity, the result will be 
that the rest of the world will suffer due to the pressure imposed by 
illegitimate intellectual rights claims carelessly endorsed by the US 
government.

As I'm sure many others will point out, "reasonable, 
nondiscriminatory" fees will rule out open source implementations. 
Imagine where the world would be if the processes of email delivery 
or HTTP were subject to "reasonable, nondiscriminatory" fees: no 
sendmail freeware, no Apache freeware. SQL? No Postgres, no mSQL, no 
MySQL. And so on. The computing world would be a different, 
substantially more impoverished place.

Granting standard status to a patent-encumbered specification is 
tatamount to granting the patent holder a license to print money. The 
W3C will demean itself in the eyes of millions if it does this. 
VoiceXML 2.0 should not be a standard for that reason alone.

Samuel Bayer

Received on Wednesday, 24 October 2001 13:58:48 UTC