WS-TXM split and BTP

ws-caf message

I was on the call when the TC confirmed the F2F decision to split WS-TXM into three parts, one for each of the current three protocols. I supported this move for the following reasons:

a) I felt it would help us to consider the requirements for, and the merits of each in turn. I always felt that where we ended up might be quite different from where we started of from, or it might not be so very different - open discussion.

b) on a more detailed technical level, TXM currently consists of two levels of XML schema. There is a common layer, then a separate schema for each of the three protocols. Looking at this I felt it was unhelpful and that the bits of common schema that were really common and useful (possibly a null set! - but maybe not) should be forced down into CF and / or Context, and anything that remains that is required by one of the transaction protocols forced up into that schema. (By the way, we really do need to find another name for CF, such as WS-Reg or WS-Registration as the one thing it is not is a coordination framework!)

c) Having got N protocol documents mapping onto CF/Context making that N+1 by adding in BTP to the family would not look at all odd. In fact it could be considered a quite logical step. Another 'by the way': It seems to me that BTP could be added to the WS-CAF family in one of two ways. We could produce a new complete version that mapped onto CF/Context or we could produce a 'profile' document that referenced the current BTP specification and specified how to use it (and what was not used) and how to map it to SOAP / CF/ Context in this case - in other words a deltas document.

Having got to this stage, I find the previous threads on this topic somewhat disturbing. However, whilst I found Mark's brief summary of the characteristics of WS-ACID as it currently stands that kicked these threads of useful, I have to agree in essence with Alastair that perhaps we should now take a step back and say what are the requirements that WS-ACID and indeed the whole 'WS-TXM' suite are trying to meet?

If end up back again at four protocols to meet all the different requirements then I for one would be reasonable happy. If we are able to produce a single, converged protocol that meets the requirements then I would be even happier!

Best Regards Tony

A M Fletcher

Tel: +44 (0) 1473 729537 Mobile: +44 (0) 7801 948219

tony.fletcher@choreology.com amfletcher@iee.org (also tony_fletcher@btopenworld.com)