RAND

I have been monitoring the discussions since the last time I commented, some 10 days ago, and I have an additional point to make.  This is under the assumption that the W3C has any business producing a patent-encumbered standard in the first place (which I dispute).

The concept of software patents is the same as that for a regular invention.  If you invent it, you get exclusive rights to that invention for a fixed length of time (according to others, this is 17 years in the U.S.).

The implementation of patents, however, is broken, very much so in the United States.  The reasons they're broken goes to he very heart of the matter of patents: legitimacy and timeliness being the biggest.

First, what is legitimately patentable?  If it is patentable, it has to be something that would be considered novel to a reasonable practitioner of the art and for which no prior art exists.  The US Patent Office granted a patent to someone for the process of prepending 2 digits to indicate century on a 2 digit year.  Besides being an obvious solution to any programmer worth his salt, there was a substantial amount of prior art.  For example, I was building applications using this technique two years before the "inventor" claimed to have invented it.  Obviously, the legitimacy of software patents relies on the standards being applied fairly and correctly.

Second, what is a fair standard for timeliness when standards and versions of standards for software come and go?  For instance, a group "invents" a software device and starts a patent application.  The standards body issues a standard that would be impacted by the patent if granted.  What happens if the patent isn't granted versus what happens if it is?  What happens if the standard gets replaced by a new (non-encumbered) version before the patent is granted?  The resulting confusion would be extreme.  

The bottom line is that by getting into the business of working around and accepting software patents as part of its standards-making process, the W3C is creating a situation where unfair patents can be granted legitimacy.  This is above and beyond my concerns that accepting RAND licensing terms endorses a process which will cause an unacceptably high barrier to entry for independent developers and small shops, and that RAND is incompatible with many open source licenses, thus leading these developers into a brick wall.

Regards,
Robert Dean

Received on Friday, 12 October 2001 13:17:01 UTC