I agree about the name - Google "SML" and you don't see the Service Modeling Language. FWIW. Dave. -----Original Message----- From: ogsa-wg-bounces@ogf.org [mailto:ogsa-wg-bounces@ogf.org] On Behalf Of Donal K. Fellows Sent: 15 January 2007 14:08 To: ogsa-wg Cc: Heather Kreger Subject: Re: [ogsa-wg] Service Modeling Language (SML) educational section Hiro Kishimoto wrote:

In July 2006, Microsoft, IBM, HP, Intel and several other companies have published "service modeling language (SML)." This is a XML based language specification and expected to be used to model complex IT services and systems, including their structure, constraints, policies, and best practices. SML is based on a profile on XML Schema and Schematron.

http://www.microsoft.com/windowsserversystem/dsi/serviceml.mspx

This spec is still proprietary and authors are going to submit it to one of SDOs maybe early this year (Q1).

Unfortunately, I will not be attending the telecon (it's a really terrible time for me!) so I'm going to have to do my comments here. The first point of note is that the name is very unfortunate, since it clashes with Standard ML. At this stage (i.e. prior to submission to a standards body) it is still reasonably easy to change the name, and explaining the change would be pretty easy. Changing to SModL or something like that would work, and not require major revisions to any documents. (This is just my opinion, but I admit to hating overloading of TLAs, since it causes trouble down the line, time and time again.) The other point of note is that there are no examples that describe how a service might actually be modelled. Given the stated purpose of the language, this seems a curious omission. It's also not clear to me whether stateful services would be modellable (e.g. so that if, in the examples from the SML spec, there was an "Enrol" operation, then it would be possible to precisely characterize the fact that success will only happen if the state of the student is changed) but that might well be a misunderstanding on my part. I was also going to ask about how to validate time-dependent properties of the service, but then I recalled that WS-ResourceLifetime makes the required information available and so reasoning there requires no new elements, assuming that the general "service reasoning" point raised above is appropriately tackled. Donal Fellows (not having had time to properly analyse the SML specs.) -- ogsa-wg mailing list ogsa-wg@ogf.org http://www.ogf.org/mailman/listinfo/ogsa-wg