RealTime IT News

SOAP 1.2 Passage Snagged By IP Issues

As the World Wide Web Consortium (W3C) churns along in its quest to ratify the SOAP 1.2 specification, a potential stumbling block has reared its head in the last week: two firms in the W3C XML Protocol Working Group said they may have patents covering some of the intellectual property covered in the proposed specification.

WebMethods and Epcientric, recently purchased by Vignette, have come forward to acknowledge that they were looking into whether or not SOAP 1.2. features technological aspects they own copyrights to.

Thursday Epicentric backed off its patent claim, leaving webMethods as the sole dissenting firm.

"Epicentric has no patents that implicate the SOAP 1.2 specification and, even if it did, would have no intention of charging royalties for the use of SOAP 1.2," Epicentric said in a company statement. "Epicentric is in the process of amending its W3C License Type from RAND (reasonable and non-discriminatory), which was an oversight, to RF (royalty free)."

WebMethods refused comment on the matter Friday.

If it is determined the spec does cover IP aspects associated with patents from the firms, webMethods would be able to charge royalties under Reasonable and Non-Discriminatory terms (RAND), which maintains that the application of royalty fees is at the discretion of the patent holder. The problem here is that the W3C may not include RAND technologies in its specs. IBM does not hold a license in this working group, while Microsoft and Intel have agreed to royalty-free status.

The dilemma comes at an unfortunate time for the W3C XML Protocol Working Group, whose members have whittled down to 11 from 393 the number of issues raised by industry participants around the development of SOAP 1.2 since June 2001, and are no doubt anxious to push the spec forward.

W3C spokesperson Janet Daly told internetnews.com neither firm has officially laid claim to a patent number, so "the [W3C] working group is not in a position to assess what technology they're asserting." Daly said webMethods has already stated it would not overlook the patent overlap.

Where to from here?
Where does that leave SOAP 1.2? Hanging, for now. The snag halts the consortium from propelling SOAP 1.2 into the "proposed recommendation" stage, where specs are ratified, until a resolution is made. That could take a while, if history is nay indication.

This dilemma is neither endemic (remember BT's Hyperlink patent claim in which the firm tried to incur royalties for all Web hyperlinks?) nor new to the W3C. In 1998, a company called Intermind came forward to challenge the W3C on its ratification of the Platform for Privacy Preferences Project (P3P), a framework for products and practices that lets Web users control the amount of personal information they share on Web sites they visit. P3P 1.0 was ratified in April 2002, but it endured its share of strife prior to that.

Ultimately, the patent group was able to prove that that it would be possible to implement P3P without infringing the patent. But not without a stultifying result.

W3C's Daly said Intermind's P3P patent claim put a "chilling effect on whether or not anyone would pick up the standards for 12-18 months."

"They issued a call for prior art and came forward with a patent number," Daly recalled. "We hired IP attorneys and formed a patent group to look into the matter."

Daly said this is one possible recourse if webMethods can come up with patents that are similar to some of the IP in the SOAP 1.2 spec. The likelihood that standard would be anything more than delayed is slim, she said.

RAND, the very regulation that permits webMethods and Epicentric to have just cause to come forward, ignited a firestorm of feelings last year, with developers clamoring to cry that patent policies are exactly what the 3C isn't about. The working group backed off from its RAND efforts in February.

Analyst sound off
Jason Bloomberg, a senior analyst for XML and Web services research firm Zapthink, doesn't think the companies are out to stymie the standard -- just a little power posturing.

"In my opinion, any company in Epicentric's and webMethod's position would realize that asserting an IP claim on part of a proposed standard will prevent that particular IP from making its way into the standard. If that IP is critical to the development of the standard, then the standard might be stymied. If not, then the standards body would have to find another approach to solving the particular issue. In either case, nobody would pay that vendor any licensing fees so that they could use the standard. Since it's not in the vendor's best interests to stymie the standard, and there's no revenue to be gained by deflecting the standard body, then why would a vendor actually take such an action?

So, Bloomberg concludes: "So what I see happening is a chess move, where white is placing black's king in check. White doesn't expect to take black's king -- in fact, that can't happen -- but rather, by presenting a threat white is forcing black to take action. Just so with vendors like Epicentric and webMethods. By threatening to exercise IP rights, they are hoping to force standards bodies to take actions that will respect the IP rights they do want to leverage -- not to prevent the standard from developing or to take a cut once it does."