After a surprise last-minute request to join the W3C's Web Services Choreography Working Group, Microsoft has just as abruptly withdrawn its two representatives, reports InfoWorld. The official line is that Microsoft wanted to know what the working group was going to be doing in relation to an aspect of choreography known as contract language. Its representatives resigned after establishing that the direction on contract language was not what Microsoft was interested in.
That sounds like a smokescreen for the real reason. Most likely, Microsoft decided to check out the new group to see how much of a threat it posed to Microsoft's own plans to secure dominance for the BPEL4WS specification that it has developed in concert with IBM. I don't suppose I'm alone in noting that BPEL4WS hasn't been submitted to any standards body, almost eight months now since it was first announced. Even then, IBM's WSFL, which was folded into BPEL, had been similarly sitting in pre-submission limbo for more than a year.
I wouldn't be surprised if IBM and Microsoft are hoping to secure market momentum for BPEL in hope that it will have become a de facto standard by the time they put it into any formal standards process. That will save them the bother of having to incorporate any pesky modifications suggested by people who've put a lot of energy already into alternative standards such as BPMI.org's BPML, and the OASIS BTP specification.
But then perhaps the W3C group is as irrelevant as Microsoft seems to believe, since it has chosen to focus on a third specification, WSCI. When you have three separate specifications being promoted by three separate standards bodies, you run the risk of political rivalries between those standards bodies overshadowing the interests of the wider industry never mind those of customers, who of course tend to come in third in the pecking order after a) standards bodies and b) vendors. The whole landscape of choregraphy and orchestration standardization is a mess, and Microsoft's quick waltz with the W3C choreography group is unlikely to be the last surprise move in this soap opera.
posted by Phil Wainewright 2:55 PM (GMT) | comments | link
Big guns go mobile
One of the most striking aspects of the latest mobile computing announcements is the focus on loosely coupled architectures. The cellular structure both of mobile networks and of 802.11b WiFi networking is one that assumes links are going to be lost and remade. Applications that run across that kind of network must also be loosely coupled, so that a temporarily lost connection doesn't undo partially completed processes. Therefore, web services and mobile computing go hand-in-hand they both operate within a distributed, componentized, service-oriented architecture.
A succession of announcements from the big guns of the industry has sent a clear message about the cornerstone technologies of this emerging architecture. When Microsoft, Intel and Cisco all focus on the same thing, then you know it's significant:
Bill Gates on Wednesday emphasized the importance of mobile computing for Microsoft, following on from Tuesday's launch of the .Net Compact Framework. He clearly believes web services go hand-in-glove with mobile computing: "XML Web services are a perfect fit for mobile devices which allow code on a device to go and get services remotely."
Last week, Intel effectively sealed the future of WiFi when it introduced the Centrino brand, which puts WiFi into laptops as standard, accompanied by a multi-million dollar marketing campaign to encourage the spread of WiFi access points both in public spaces and within corporations.
The move left Cisco scrambling to play catch-up, announcing yesterday its acquisition of Linksys, a company characterized by most news reports of the deal as a home networking specialist, but which Cisco bought primarily because of its WiFi product line.
In a further interesting development, Nokia today unveiled plans to converge its handsets with WiFi, reports InfoWorld: "Nokia executives have revealed plans to deliver high-end cellular handsets that seamlessly switch between 802.11b and public carrier networks in an effort to transform handsets into terminals capable of consuming composite applications."
Sales of web services development tools slipped last year and won't recover till 2004, according to a new report from Aberdeen Group. The report blames the slowdown on "over-hyped and under-performing products," but that's a little unfair on the vendors. In the first flush of enthusiasm for web services, developers were going to buy anything they could lay their hands on, fully aware that it was going to be roughly-hewn, unproven, first-generation technology. It was the market that was in a state of hype, not the products or the vendors.
Now that developer ardor has cooled, so have sales. This entirely predictable turn of events merely signifies that developers have already passed the 'peak of inflated expectations' foreseen in Gartner's hype cycle. This is followed by a period when vendors hone their offerings and developers refine their understanding of how best to apply the technology, resulting at last in a flourishing of productive use but not before a backlash of discontent has unleashed a torrent of complaints and backbiting about the inadequacy of the earlier tools. The tone of Aberdeen's report is perfectly timed to cash in on that sense of discontent, but it's not revealing anything that should come as a surprise to anybody.
What I find more interesting is that there's still a lot of media hype and excitement about web services, even though developers are well past that stage. Earlier, this year, I noted an observation by Microsoft's Peter Bell that IT architects and business strategists may still be climbing towards the peak of hype, even while developers descend the slope of Dilbertian disillusionment that lies beyond. I would add to that thought the observation that there are probably several related hype cycles in play digital identity, orchestration and business process management to name but a few likely suspects and so the image to hold in mind is that of a straggling army of hillwalkers tackling a serried chain of peaks as they progress towards their ultimate destination, each of them convinced that they're in the leading party, and yet frustrated in their efforts to communicate directions or co-ordinates that make any sense at all to any of their colleagues.
posted by Phil Wainewright 3:35 AM (GMT) | comments | link
Assembling on-demand services to automate business, commerce, and the sharing of knowledge