* Software Patents: W3C Standards (Part 7)

The open source community has generally viewed the W3C’s decision on patents in standards as a victory.

The W3C (http://www.w3c.org/) was formed, in part, to fast-track Internet standards relating to the Web. This snippet from the W3C’s 1998 FAQ describes it best:

“The Internet Engineering Task Force (IETF) is a set of working groups which meet on the net and in person three times yearly; now under the auspices of the Internet Society. The Internet Society is a professional society.

The W3C is open in the sense that any organization may join who is prepared to fund the design and coordination teams.

The IETF is open [in] the sense that any volunteer may come along. It does not fund design.

IETF working groups provide excellent environments for brainstorming over new problems, and for critical review of proposed standard protocols. W3C Team members participate in IETF Working Groups, gaining from the wealth of ideas and the criticism. However, with no engineering budget, the IETF cannot make commitments to deadlines for products. The W3C teams do actual design work, on both specifications and reference code. The W3C has its own process for endorsement of work by its Members, as defined in the membership agreement.

W3C results may also be put on the IETF standards track, but if the IETF process of open discussion does not make rapid progress, W3C Members need not be held back.”

As a result, the HTML 4.0 specification is on the W3C website (http://www.w3.org/TR/REC-html40/) but is not an IETF RFC (http://www.rfc-editor.org/rfcxx00.html). And there is evidence to support the proposition that the IETF moves slowly (although whether it moves too slowly is debatable) as the HTTP specification is still a draft standard (http://ftp.rfc-editor.org/in-notes/rfc2616.txt).

Also, the W3C focuses on “higher level” internetworking standards (such as HTML and XML) and whether you use the OSI seven-layer model or the TCP/IP four-layer model, the higher the level, the more likely you’re dealing with software, and the lower the level, the more likely you’re dealing with hardware. For what it’s worth, some people really don’t like the OSI seven-layer model (http://www.randywanker.com/OSI/). This is a long way of saying that the W3C deals primarily with software.

On 05/20/03, the W3C published its revised patent policy (http://www.w3.org/Consortium/Patent-Policy-20030520.html), which (unlike the IETF’s current policy) rejects the use of patents in WC3 standards. The open source community has generally viewed the W3C’s decision as a victory.

Michael Tiemann, Red Hat Chief Technology Officer, summarized the history of the issue and how the decision came about (http://www.redhat.com/advice/speaks_w3c_patent.html). Tieman states:

“Seeing that the open source community could recover the notion of truly open standards that helped give birth to the World Wide Web, we should be encouraged that we can win back the freedoms that are under attack in other arenas, threatened by other organizations, and enjoined by other legislation.”

But doubts remain about the new W3C policy, such as those in the comments to a NewsForge post (http://newsvac.newsforge.com/newsvac/03/05/24/1618229.shtml):

“When the draft of the current patent policy was aired, concerns were raised over the Section 3 clause ‘3. may be limited to implementations of the Recommendation […]’, which allows the patent license to impose a field-of-use restriction. This is incompatible with many free software licenses, most notably the GPL.

http://www.fsf.org/philosophy/w3c-patent.html

Essentially, a patent license that only allows the technology to be implemented for a particular use (a field-of-use restriction) cannot be implemented in software that is licensed under terms specifically disclaiming field-of-use restrictions, as the GPL does. W3C technologies encumbered by such field-of-use restrictions would be incompatible with free software.”

Bruce Perens is also still skeptical (http://news.com.com/2100-1013-996351.html):

“We need to be clear that making a patent available for (free) use in a standard doesn’t mean the patent holder waives all royalties…. The royalty-free terms at W3C are for implementation of the standard only, not for any other purpose — even in the same program. That gives the patent holders lots of room to make money.”

In summary, the open source community, in general, appears to be more pleased with the W3C’s patent policy than with the IETF’s. The IETF plods along and generates good Standards that include patent technology when it makes sense to do so. The W3C’s faster-is-better approach excludes patents and is consistent with the reasons the W3C was created in the first place. It should come as no surprise that the IETF’s policies include patents and the W3C’s do not.