The W3C today denied that it has misled the web community on the status of its controversial RAND license, but confirmed to us that RAND was effectively already in "beta".
In its FAQ in support of the pay-to-play web standard RAND licensing policy, the W3C emphasizes "NOTE: The W3C is not presently aware of any non-free patent that is essential to any existing W3C recommendation." And Danny Weitzner reiterated this to The Register yesterday, telling us, "...we observed some general trends, and concluded that we should at least think about RAND."
However Linux kernel hacker Daniel Phillips, has pointed out that a draft version of the RAND license has been part of private discussions since April 5. On the SVG Patent Statements Page IBM, Apple, Kodak and Quark declare for RAND licenses. Kodak named a patent that could be an issue. Phillips today repeated his call to fork the SVG specification today.
How can we square the two viewpoints?
Janet Daly of the W3C says it's a "mischaracterization" to say that RAND is already on the table.
"We were able to test a draft and see what issues arose," she told us.
"What we've hoped to do in applying the early version of the Patent Policy Draft to existing Working Groups is identify patent issues early. Remember that there was no official declared mode for the SVG working group and it's now a RF group. Kodak said the patent they'd identified was non essential, and will help produce open source implementations."
Weitzner added that a RAND would identify earlier, rather than later, any patent issues.
"RAND is part of our environment, without being a policy," he told us today. "Remember that there's nothing in the W3C's procedures that says a specification should be either RF or RAND."
His view concurs with many opponents of RAND - that RAND licenses would most likely sink in the marketplace. "But rather than just bumbling along, people can identify the patents".
That's not enough to satisfy free software community.
We've been searching for common ground, and noticed a couple of things.
While there's probably much shared philosophical agreement that web standards don't work when there are royalties attached, or when there are submarine patents waiting in the dock, and much agreement that public opprobrium is the best way to counter punitive patents, there's little agreement on tactics: on where the best time and place is to bring these into the open, so the public can throw embalming fluids at the perpetrators.
And there's even less appreciation on how much damage RAND could do to the free software movement.
"You can't license a patent encumbered standard under the GPL," says Philips, who dismisses Kodak's concession to produce an open source implementation as too vague a promise.
"We have to do a good job on researching patents and we have to move rapidly when a submarine comes up, to embarrass the company with the patent," he says.
"But there's got to be no compromise on RAND. Only RF is acceptable - and that applies to the IETF as well." ®