Will patents drive us to REST?

by Simon St. Laurent

As IBM and Microsoft grow ever cozier, posting specifications with unpleasant intellectual property ramifications, and setting up organizations that seem quite likely intent on building "Reasonable and Non-Discriminatory" (RAND) tollbooths, maybe it's time for the rest of us to appreciate what HTTP and XML have given us, and take a REST.

Covering Web Services and the intellectual property territory they represent is growing pretty terrifying. At first I thought David Berlind's suggestions that "IBM and Microsoft plan Net takeover" were fairly ridiculous, but IBM's apparent patent bombing of ebXML (which IBM now claims to have retracted) certainly makes it more plausible, as do the intellectual property policies of the Web Services Interoperability Organization and UDDI.org. I used to worry about the W3C, but vendors seem to be heading for more controllable consortia as the W3C focuses on royalty-free specifications.

While SOAP itself seems, for the moment, to be unencumbered, WSDL, UDDI, various security add-ons, and possibly alternate transport mechanisms are all looking questionable.

REST, even apart from the ongoing debates over its merits, is pretty much built on HTTP and HTTP principles, with possibly a dose of XML. There isn't a whole lot in REST (or in XML) that is strikingly new, and hopefully there isn't anything there that is, well, patentable.

However you feel about the technical merits of Web Services or REST, it seems pretty clear that REST is a critical option that offers developers an apparently safe harbor from intellectual property claims. REST may let us get our work done, whatever the outcome of various patent claims, and may also give patent-holders a reason to ponder a more generous approach to their intellectual property.

Do patents on various aspects of Web Services affect your decisions about which technologies to use?