Decision time for WS-I members

The 166 members of the Web Services Interoperability Organization (WS-I) have just five days next week to cast their votes to decide which two companies will join the original nine founding members on its board of directors. Voting opens on Monday and closes Friday March 21st. The results are due to be announced a week later, but will doubtless be whispered widely several days in advance, such is the level of interest in the contest.

The primary aim of the election is to rectify the omission of Sun Microsystems from the original board. Sun was excluded because Microsoft (to paraphrase Groucho Marx) couldn't bear the thought of founding a club that would have Sun as a leading member. But it rapidly became clear that it just wasn't tenable for WS-I to exclude the company with ultimate responsibility for Java and the J2EE platform, so the election ruse was concocted as a face-saving route to bring Sun on board.

In principle, then, this is a fixed election in which the winning candidate is known in advance. No one should be under any illusion that the politics of industry bodies is either representative or democratic. It has far more in common with the politics of medieval Italian city states, as described in the writings of Machiavelli. But there are no guarantees that everything will go according to plan.

It all depends on how sophisticated a campaign the various contenders are pursuing. I suspect Nokia and SeeBeyond have shown their hands too late in the game to have a serious chance of winning, but Cape Clear, WebMethods and Verisign have all had time to run a credible and thorough campaign. Since everyone expects Sun to win, many members may feel they can afford to cast their votes in favor of other contenders, but a strong showing by other runners could lead to an embarrassing upset:

1. A total eclipse, in which Sun gets pushed into third place, is likely only if Sun itself has taken its election so much for granted that it isn't doing any campaigning at all. Unfortunately, Sun's track record with industry bodies suggests that it is capable of committing an error of judgement of this magnitude. Hopefully wiser counsel has prevailed.


2. A partial eclipse, in which Sun comes second, is a stronger possibility. If a large number of members all lean towards one particular contender, and if that contender converts that support into votes by diligent canvassing and a systematic campaign, then it could end up running Sun surprisingly close for first place.

On balance, I think Sun is still going to come in first, but the result for the other place could have a significant impact on how WS-I develops in the future. My money is on Cape Clear, which I think makes a persuasive case for representing the smaller members of WS-I, and which is well connected within the industry. WS-I's current board of Accenture, BEA, Fujitsu, HP, IBM, Intel, Microsoft, Oracle and SAP looks a bit too much like the establishment old guard for my liking. Adding Cape Clear would spice things up a bit more to my taste.

But whatever the outcome of the election, WS-I's work is already progressing and gets a good write-up this week in an article by Rich Salz at XML.com. Introducing WS-I and the Basic Profile is a very clear exposition of what WS-I does and doesn't aim to do, and gives a useful explanation of the Basic Profile drawn up by WS-I: "It is, frankly, an amazing document. Web service developers who are concerned about interoperability should get a copy, read it, and send in feedback," he writes.
posted by Phil Wainewright 10:51 AM (GMT) | comments | link

Why buy CrossWeave?

AmberPoint's VP of marketing says it's "not interested" in building composite applications, even though it just bought CrossWeave, a company that specialized in doing just that. Why then did AmberPoint buy CrossWeave?

Part of the explanation is that CrossWeave was, quite literally, just begging to be bought. According to a CRN report of the deal, CrossWeave "was struggling to remain afloat." Its main VC backers, Norwest Venture Partners and Sutter Hill Partners, will have smoothed introductions to Amberpoint, since both are also key investors in the web services management startup. And with both companies based in Oakland, there were few logistical barriers.

One asset that will have interested AmberPoint is CrossWeave's customer base and prospect list. In the week the deal was done, we published an article here, Too few web services to manage, reporting on the shortage of customers for web service management tools: "Few web services implementations are seen as complex enough to require management."

As one of the earliest pioneers in assembling composite applications, CrossWeave's customers comprise some of the rare exceptions, and it is likely to have built up a database of similar prospects that AmberPoint will be keen to harvest. Right from its launch in June 2001, CrossWeave was advocating the use of web technologies to hook together functionality out of different applications, using terminology and turns of speech that sound familiar today in the context of web services, but were well ahead of their time back then.

I interviewed Sean Fitts, the company's founding CTO, for an article back in July 2001. "Because you've got a common internet infrastructure and common protocols, [the web environment] gives you a way to integrate applications in different ways than before," he told me. CrossWeave's message was all about using web technologies to link functionality across existing applications without needing to change the original code: "We're helping customers to unlock value from things they've already invested in," Fitts said. My subtitle for the article was "loosely coupled integration," which is something of a landmark in itself, because as far as I'm aware, it was my first published use of the phrase.

This loosely coupled integration is exactly what web services management vendors are now directing their attentions towards, we concluded in Too few web services to manage: "Web services management isn't just about managing brand new web services applications; it's about managing the interactions that web services enable between many different forms of applications, within a service-oriented architecture (SOA)."

So Ed Horst, AmberPoint's VP of marketing, was being a little ingenuous when he told InfoWorld this week: ""[CrossWeave's] purpose in life was to build composite applications, where you take pieces of an existing application and combine them into new applications. We're not interested in doing that."

Although AmberPoint isn't interested in building composite applications itself, it is most definitely interested in having customers and partners who build composite applications, because they will become ideal prospects for its management products. We may well therefore see elements of CrossWeave's software emerging as a toolkit or even as licensed technology that AmberPoint's vendor and systems integrator partners can use to help them create composite applications. In buying CrossWeave, AmberPoint, like its peers, is investing in encouraging the kind of service-oriented integration that will lift demand for its management capabilities.
posted by Phil Wainewright 2:50 AM (GMT) | comments | link