Event Metamodel and Profile OMG standard -draft RFP released

[PVincentTIBCO]

See http://www.omg.org/cgi-bin/doc?ad/2007-11-01 for more information. Kudos to PJ of Rhysome for this work.

The standard basically asks for a 1st-class modelling entity for "event" that can be used across event processing paradigms (eg BPDM/BPMN for BPM, EDOC, etc etc) at the modeling level, with options for any standardised event processing modeling methods too ( ). Possibly a group of submitters will meet up to discuss a joint response at the next OMG meeting (December, Burlingame) when the RFP is voted to be published (or not).

Cheers
PaulVincent (for standards in EPTS)

PS: Wonder if Standards should have its own category on ths forum?

[DLuckham]

Paul, I took a look at this doc. Can't claim to have read it in detail. My understanding is that this is a Request for Proposals on models of events. It seemed awfully complicated. I was wondering if (1) the doc defines "event" anywhere. Is it in line with the Glossary definition? (2) Is an event model from OMG supposed to be a super class (or template) of existing event definitions such as in SOAP (I'm presuming there is one in SOAP, not having read that protocol in detail either)?
And does it have to be in UML? Is there no other formalism that would be appropriate? - Cheers David

[Tim Bass]

Dear Paul and David,

I read the OMB document in detail quite some time ago.

My impression was the document could easily be used by OMG group members to gather competitive intelligence on the event processing offerings and plans by other software vendors.

In particular, at least one of the vendors leading the effort was a small CEP software vendor.

This (intelligence gathering) is one of the problems with these types of documents and "standards efforts" at this early stage of the market.

Yours faithfully, Tim

[Tim Bass]

Dear Paul,

I certainly agree with you that there should be a forum category for standards, something like:

Event Processing Standards and Related Activities

(or some similar title)

Yours faithfully, Tim

[PVincentTIBCO]

Paul, I took a look at this doc. Can't claim to have read it in detail. My understanding is that this is a Request for Proposals on models of events. It seemed awfully complicated. I was wondering if (1) the doc defines "event" anywhere. Is it in line with the Glossary definition? (2) Is an event model from OMG supposed to be a super class (or template) of existing event definitions such as in SOAP (I'm presuming there is one in SOAP, not having read that protocol in detail either)?
And does it have to be in UML? Is there no other formalism that would be appropriate? - Cheers David

Hi David - yes, and this is the uber-simplified version! Apart from allowing submitters the headroom of proposing Event Processing technique modeling, my view is that the main role of EMP is to make "event" a first class citizen in the UML world, for simple event processing (BPM etc) as well as CEP. If EMP achieves just that, with a standard model definition of event, it will be worthwhile. Other submitters may have different opinions of course... I will ask them to consider posting here.
Cheers,
Paul

[PVincentTIBCO]

...My impression was the document could easily be used by OMG group members to gather competitive intelligence on the event processing offerings and plans by other software vendors.

In particular, at least one of the vendors leading the effort was a small CEP software vendor.

This (intelligence gathering) is one of the problems with these types of documents and "standards efforts" at this early stage of the market.

Yours faithfully, Tim

Hi Tim - you are right, although the Request For Info stage which happened earlier is meant to do the intelligence gathering! I'm not sure who responded to that (will need to check on the OMG site). However, having a single model representation of event is probably not too much of a competitive issue between vendors.

Earlier drafts went on about standardizing EP techniques, which of course is a huge subject and not likely to be successful IMHO.

Note that there is certainly a tendancy in standards bodies for small companies to put effort into a standard to their own market advantage. A few of the standards coming out of OMG are from "minority views". So you are absolutely right to be cautious about this.

Hopefully the involvement of the EPTS will ensure an agreeable EMP standard that is compatible with all vendors...

Cheers,
Paul

[Tim Bass]

Note that there is certainly a tendancy in standards bodies for small companies to put effort into a standard to their own market advantage. A few of the standards coming out of OMG are from "minority views". So you are absolutely right to be cautious about this.

Hopefully the involvement of the EPTS will ensure an agreeable EMP standard that is compatible with all vendors...

Cheers,
Paul

Hi Paul,

The document I saw went far beyond questions about standardization efforts. As I recall (it has been quite a while) the RFI was many pages of gathering information, only a subset directly related to standards (if we are talking about the same document)!

Another problem, as I recall, was that the raw data was (is) not being made publicly available - however, if you are a OMG-insider vendor, you have full access to the data.

Cheers,

Yours faithfully, Tim

[PVincentTIBCO]

...The document I saw went far beyond questions about standardization efforts. As I recall (it has been quite a while) the RFI was many pages of gathering information, only a subset directly related to standards (if we are talking about the same document)!

Another problem, as I recall, was that the raw data was (is) not being made publicly available - however, if you are a OMG-insider vendor, you have full access to the data.

Cheers,

Yours faithfully, Tim

Tim - Some OMG info is public (http://soa.omg.org/SOA-docs/EDA-Standards.htm ?). But I couldn't find any RFI responses myself so maybe everyone had the same concern as you!
Cheers,
Paul

[PVincentTIBCO]

PS: an interesting issue for standards bodies is that CEP covers so many domains.

http://www.omg.org/news/meetings/worksh ... evised.pdf for example was presented to the Real-Time group at OMG this year. But EMP (and Opher's previous presentation) was made to the SOA group...

[PVincentTIBCO]

... But I couldn't find any RFI responses myself ...

Bob at Rhysome has politely corrected my myopia with http://www.omg.org/techprocess/meetings ... p_RFI.html being the (OMG member-only) link to the RFI responses. But of course these are only justifications for interest in the EMP RequestForProposals.

There may be a meeting at OMG Burlingame on EMP submissions / teaming up. If so, I will be wearing the EPTS (as in reporting back to) and TIBCO (as in assessing approach and team to join) hats.