Welcome Web Services Activity

January 30, 2002

Edd Dumbill

This week the World Wide Web Consortium announced the formation of a Web Services Activity. Within the W3C, "Activity" is the name given to an ongoing focus of development encompassing one or more Working Groups. Until this time, the W3C's only participation in the web services world was through the XML Protocol Working Group, which is essentially tidying up SOAP.

Since the formation of the XML Protocol Working Group, several companies followed the example of the SOAP team and joined together in ad-hoc groupings to develop the complementary machinery needed to make SOAP work with their programming environments. One technology devised this way was the Web Services Description Language, WSDL, which has become closely intertwined with the use of SOAP. There are pitfalls with trying to standardize something that's brand new, and WSDL has come in for some criticism, in the same way as SOAP did.'s own web services columnists recently made a plea for a W3C Working Group to take on WSDL. With the announcement of the new W3C Activity, they got their wish.

The Activity comprises three working groups: XML Protocol (transferred from the XML Activity), Web Services Description, and Web Services Architecture. Additionally, there is a Coordination Group to ensure coherence among the working groups. It is the Architecture group which is the most interesting of these, as it embodies the essential advantage of this development work taking place within the W3C rather than elsewhere. The role of the Architecture group is, unsurprisingly, to design the overall architecture of the clutch of technologies that will constitute web services. According to its charter, the group has the following goals:

  • The architecture described must be modular: a set of technologies will need to be described to address individual functionalities identified by the architecture document.

  • The set of technologies identified must be based on XML.

  • The Web services architecture will need to be cleanly integrated in the Web architecture: Web services should be addressable resources, results of operations which do not have side-effects should be cacheable, etc.

  • The architecture designed by the Working Group must be platform independent, must not preclude any programming model nor assume any particular mode of communication between peers.

  • Focus must be put on simplicity, modularity and decentralization.

  • The framework proposed must support the kind of extensibility actually seen on the Web: disparity of document formats and protocols used to communicate, mixing of XML vocabularies using XML namespaces, development of solutions in a distributed environment without a central authority, etc. In particular, it must support distributed extensibility, without third party agreement, where the communicating parties do not have a priori knowledge of each other.

What is particularly welcome about these goals is their sympathy with existing Web infrastructure. Often when new technologies are devised independently they build on an ignorance of existing standards, storing up trouble for the future as the technology becomes adopted. The Web is an existing application and all new additions need to recognize that. If the Architecture WG achieves its aims of Web integration and distributed extensibility it will go a long way to ensuring the continued openness and relative "un-brokenness" of the Web: this is vital in an area that is such a playground for companies' product strategies.

The move of the XML Protocol WG into a more defined area away from the thrust of XML development is also welcome. It may enable an increased focus at the W3C on issues of more foundational importance in XML. The Protocol WG's ridiculously huge membership of over eighty reflected more the urge among companies to be associated with the latest trend than a desire to work on the technology itself. Giving web services its own Activity helps contextualize SOAP and friends, and together with the slump in the economy, will hopefully put an end to the trend of stupidly inflated working groups.

However keen the participants may be, standards take, by implication of their definition, time to produce. When the issue at hand is something so potentially far-reaching as web services, extra care must be taken. The reasons that some companies are frustrated by the "slow" development of W3C standards have more to do with marketing and product concerns than technical ones. Hence this is another reason why the W3C taking on web services is a good thing: an independent rate-limiter that increases the probability of something useful emerging from the other end of the process. For an example of how things can go wrong, look at UDDI: an initiative more motivated by marketing and corporate strategic concerns than technical ones, not especially versed in the foundations of the Web, and something that hasn't really produced anything of real use. At any rate, UDDI needs to wait for SOAP and WSDL to be be finalized before it even has a decent use-case.

Some may view the new Activity as a W3C land-grab. Early on in the SOAP days it was thought that the IETF was a better place to pursue such work. Nevertheless, right now the W3C is the only consortium in town that seems to work in a way that both satisfies corporate participants and is capable of producing something technically credible. W3C is already working with IETF on SOAP and pledges in its new activity statement to liase with groups such as ebXML and OMG.

In summary, the new W3C Web Services Activity is very welcome. Its aims are laudable indeed: the big question is whether they can be achieved. I dearly hope that in two years' time I can write that the goals of simplicity, modularity, extensibility, and integration with the Web's infrastructure have been attained.

You Know You Are ... When

I am reluctant to introduce any more gimmicks into the XML community, considering the recent deluge of XML-related limericks and haiku. However, I thought this worth sharing: Micah Dubinko, contributor and XForms expert, has recently been deeply into the XML developers' list XML-DEV. He came up with a list of "top ten signs you're spending too much time on XML-DEV," featured below.

10. You start off your message with "In our last argument, you said..."
9. Added text to quote ratio < 1/100
8. You actually read every message
7. You email your wife to say you're coming home early, and start the subject line with "ANN:..."
6. Visions of angle brackets dance in your head
5. Your sig refers to XML--in all four lines of it (sorry, Simon!)
4. You've replied in the same thread 10 times--to make a single point
3. You can no longer speak an entire sentence without the word "semantics" in it somewhere
2. You've chosen sides on "RDF vs. Topic Maps"
And the number one sign you've been spending too much time on XML-DEV:
1. You don't need to read XML-Deviant every Wednesday.


Also in <taglines/>

XML 2004: After Declaring Victory, What's Next?

An Old New Thing

Moving On, But Not So Far

XML at Five

Whither Web Services?

  • Readers will probably be familiar with tiresome unsolicited emails from anonymous email accounts inviting them to check out sordid and unsavory web sites. Indeed, it is said by some that the operators of such licentious sites lead the way in Internet technology. One XML vendor certainly seems to think so, as I found when I recently received a breathless missive from a Hotmail account. The writer, purporting to be an excited member of the public, invited me to check out a "really ground breaking ... solution to XML in an enterprise environment" they'd come across. My correspondent lauded the product as "filling a huge gap in the list of available XML applications." I won't embarrass the vendor concerned in these pages but I hope, dear Bradley, you find a career that suits you better than marketing.

  • As if the language isn't being ruined enough by such deplorable tendencies as using the word "enterprise" as an adjective, XML acronym factories (otherwise known as standards organization committees) are doing their best to further pollute our lexicon. The fascination with affiliating with a cause through acronyms is as strong as ever, with the emergent trope of "WS??" for anyone wanting to attach to the web services bandwagon. Readers should not confuse acronymity for activity, however, as demonstrated by an announcement that reached me this week. OASIS happily announced that the first, undoubtedly dynamic, action of the brand new OASIS WSCM TC was ... to change its name to the OASIS WSIA TC. With so many of these groups floating around, it can be dangerous to assume anyone really cares.

  • One of the more curious discoveries I made at the end of last year is that not everyone involved with the development of W3C's XML Query Language is terribly happy about it. I say curious, as the impression I have of this group is of a focused activity likely to lead to a sound technology. As yet, nobody has provided me with a solid reason for their unhappiness, so one must assume the normal machinations of consortium politics. It is most peculiar, however, that at the upcoming W3C Technical Plenary there is only one working group refusing observers -- something considered rather impolite at a plenary event -- XML Query. They are clearly expecting a fight!