The W3C works with developers, software makers and others to come up with standards for the Web, which can then be used by just about anyone to build Web software, free of charge. To date, either those standards have not been based on patented technology, or the holders of patents have chosen to not enforce patents in order that the standards be widely adopted.
But a new proposal may open a few cracks in that wall, allowing companies to enforce patents based on those technologies and to potentially charge a royalty fee to developers who use them.
The W3C's Patent Policy Framework, more commonly referred to as the "reasonable and non-discriminatory" (RAND) licensing proposal, acknowledges a central conflict to the standardization process: Companies that spend serious time and effort coming up with the technology behind the standards may be reluctant to simply give away the rights to what they consider their intellectual property.
That proposal, written by W3C members representing large technology companies such as Microsoft, Apple Computer, Hewlett-Packard and Philips Electronics, has ignited a firestorm of controversy among developers, some of whom claim it could mean the end of innovation for the Web and could lead to undue influence by commercial companies over the standards process.
More specifically, some developers believe the proposal calls into question the W3C's role as the arbiter of standards related to the Web. A final decision on the policy is expected from the W3C by February 2002.
Calls to Microsoft and other large technology companies were not immediately returned. A W3C representative did not immediately respond to a request for comment.
"This proposal would effectively ban open-source interpretations of these standards," said Bruce Perens, who helped write the Open Source Definition in 1998. "We don't have royalties in open-source software. The W3C can potentially marginalize itself with the RAND proposal."
Mike Todd, president of the Internet Society's Los Angeles chapter, said the RAND proposal would "create a situation where users would get used to using something that contains these sleeping copyrights, and then if they are activated, they will cause chaos."
Todd suggests that if the W3C adopts the proposal, users of W3C standards containing copyrights should be advised of the specific copyrighted code, so "people are aware that there is a certain aspect (of that standard) that is copyrighted, but if you don't use it, you are in the clear."
The W3C was founded in 1994 to "lead the World Wide Web to its full potential by developing common protocols that promote its evolution and ensure its interoperability," according to the organization's Web site. More than 500 organizations are members of the W3C, which has developed more than 35 technical specifications behind the Web, such as HTTP, XML and HTML.
In the early days of the Web, the W3C set direction on many technologies key to the Web's adoption. But as the Web has become more mainstream, and more usable, the W3C's work has become more complex and arcane, said one analyst.
"There is a sense that the (W3C) is becoming a little too academic and out of the mainstream and their work too esoteric," said Uttam Narsu, an analyst with Giga Information Group.
The RAND proposal may be in reaction to that increased complexity, said Narsu. "The W3C needs to look at streamlining the standards process by taking something that is a de facto standard" and making it a W3C-recommended standard.
"There is considerable difficulty to come up with a standard that does not infringe on a patent," Narsu said. While the W3C in the past recommended standards that are patent-free, that doesn't mean those technologies were always the best way to solve a given problem.
"It may be that the adopted standard is the second or third choice, because the preferred technology was patented," Narsu said.
Other standards bodies already adopt standards based on proprietary technology. And patents are not unknown to the world of standards bodies; the Joint Electronic Devices Engineering Council, or JEDEC, for instance, permits companies to submit patents for adoption by the group and to collect royalties from members.
While the group prefers to adopt free patents, it will adopt for-fee patents as well. JEDEC states that members have to disclose pending patents--the organization can't adopt a standard with an undisclosed patent--but it can adopt a patent into a standard if disclosed. The patent owner can also choose to license it for free or charge everyone an equal royalty.
Narsu said typically there is pressure on the patent holder from the community of users surrounding standards bodies to loosen up patent requirements, which many companies do.
Specifically, the RAND proposal would require:
That W3C working groups spell out the licensing terms for a proposal along with the technical requirements in its charter.
All W3C Members to disclose any patent claims they know of that may be essential to a recommendation. Members whose contributions become the basis for working group efforts would have an additional obligation to disclose relevant patent claims and licensing conditions at the time of their submission.
All W3C Members to make a legally binding commitment to license patent claims essential for implementing a W3C recommendation on RAND terms. If they're not willing to license particular technology on RAND terms, they must opt out of specific patent claims they hold, normally within 60 days after the publication of the last-call working draft.
Some of the controversy involves not just the proposal itself, but the timing. The period for public comment on the proposal expired Sept. 30. Although the proposal was published to the W3C's Wet site on Aug. 16, and news about it was posted on the W3C's site on Aug. 20, most of the comments indicate that people were not aware of the proposal until this past weekend. Many are calling for an extension of the public review and comment period. A W3C representative said there "has been discussion" about extending the comment period.
But once the proposal was publicized through postings on open source-friendly sites including Linux Today and Slashdot.org, the criticism was fast and furious.
"A bad policy," "Just say NO!" and "RAND is WRONG," read typical subject lines of the comments that were submitted to the W3C.
Notable open-source proponents, including Free Software Foundation President Richard Stallman, have urged the W3C to declare that all important standards must have free patent licenses.
Stallman's comment also argued that the policy may not discriminate against a specific person, but it does "discriminate against the free software community, and that makes them unreasonable."
The W3C proposal is backed by some of the largest technology makers in the industry. The working group that developed the proposal includes a who's who of technology: Microsoft, Hewlett-Packard, Philips, Apple, AT&T, IBM, ILOG, Nortel Networks, The Open Group, Reuters and Sun Microsystems, along with W3C affiliates.
And some of those companies, most notably Microsoft, have shown their disapproval for certain aspects of the open-source movement.
In June, Microsoft Chairman Bill Gates called the GNU General Public License that governs the distribution of some open-source software "Pac-man like," saying it "is impossible for a commercial company to use any of that work or build on any of that work."
One posting to a W3C newsgroup, from a person claiming to be a Microsoft employee, defended the proposal. "Patents are a critical part of our Intellectual Property systems and a key underpinning of our capitalist economy," the posting read.
The authorship of the proposal didn't go unnoticed. One commentator argued that it "has the ugly smell of a meat packer bribing the USDA."
Even if the proposal is approved, it could cause infighting among those large businesses that are requesting the right to charge royalties.
"The fighting will not just be over royalty-free (RF) versus RAND," Bruce Pezzlo, president of Plum Computer Consulting, told CNET News.com in an e-mail. "The arguments will be between each party who believes they have some patent that is related to any proposed standard. Each will jockey for a position, monitor what the others are charging as fees, and believe they too should be entitled to the same.
"The net effect will become standards will take too long to become adopted, and not widely adopted should the cost of fees become prohibitively expensive."