Workshop summary and observations (2)

To continue the summary and observations on the Web of Services Workshop last week…
Not a lot of blogging afterward yet, but here’s an entry from Pete and one from Jonathan.
It was clear that both the Web (REST) and Web services (WS-*) are being successfully used in production applications today.
The usage patterns seem to substantiate my view that Web services are more often used in enterprise IT environments that predate the Web. Most of the users who attended the workshop – and they tended to represent large, established organizations – said that they are using Web services in what I’d call mission-critical or operational applications.
It was also very clear that many of the same organizations are successfully using Web technologies in mission critical/operational applications.
Some of us – myself included – took the position that both technologies can and should co-exist, and that it would be good if the W3C could help define how this could and should happen.
I thought Noah’s paper was really great, and probably the best presentation as well. (But of course my reaction was perhaps predisposed because of my thinking that a hybrid or combined solution is what the industry needs.)
Although he is employed by IBM, Noah attended the workshop to represent the W3C’s Technical Architecture Group (TAG).
He said the TAG’s view is that people are getting value out of both Web and Web services technologies. His presentation and paper include some very good pros and cons points, and describes what I thought was a great approach to using the two in combination.
One unfortunate part is that what Noah presented relies to a large extent on parts of the specifications that aren’t well implemented, or are potentially misusable.
The SOAP 1.2 GET feature, WSDL 2.0, and WS-Addressing endpoint reference mechanisms are integral to Noah’s recommendation, but not yet widely implemented (with the exception of WS-Addressing EPRs, but this is a variation on the issue that I’ll explain later).
This is directly related to part of the discussion during the Workshop, about how the issues the Web community has with Web services relate more to how the specifications have been implemented (or not implemented, as the case may be) than with the specifications themselves.
Achieving an impact in this area may be somewhat challenging, but this may be one of the things to evaluate in the context of going forward.
The WS-A EPR issue relates to the fact that as specified, they resemble cookies inasmuch as they are intended to carry opaque data. The data in EPRs may contain identifiers, but if they do they risk “breaking” the Web in that they could be using a format other than a URI to identify an endpoint.
More later…


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s