atom feed20 messages in org.oasis-open.lists.wsbpelRE: [wsbpel] Issue - 11 - Let's get a...
FromSent OnAttachments
Danny van der RijnSep 3, 2003 10:53 am 
Satish ThatteSep 3, 2003 7:05 pm 
Chris KellerSep 4, 2003 6:49 am 
Danny van der RijnSep 4, 2003 9:29 am 
Danny van der RijnSep 4, 2003 9:35 am 
Satish ThatteSep 4, 2003 9:59 am 
David RR Webber - XML ebusinessSep 4, 2003 8:50 pm 
Chris KellerSep 5, 2003 5:26 am 
Chris KellerSep 5, 2003 5:48 am 
Monica MartinSep 5, 2003 6:27 am 
Chris KellerSep 5, 2003 6:32 am 
Monica MartinSep 5, 2003 6:40 am 
Chris KellerSep 5, 2003 6:41 am 
Danny van der RijnSep 5, 2003 10:33 am 
David RR Webber - XML ebusinessSep 5, 2003 11:54 am 
David RR Webber - XML ebusinessSep 5, 2003 11:59 am 
David RR Webber - XML ebusinessSep 5, 2003 12:03 pm 
Ugo CordaSep 5, 2003 2:38 pm 
Danny van der RijnSep 5, 2003 3:55 pm 
Francisco CurberaSep 11, 2003 6:07 pm 
Subject:RE: [wsbpel] Issue - 11 - Let's get a proposal nailed down
From:David RR Webber - XML ebusiness (Gnos@compuserve.com)
Date:Sep 4, 2003 8:50:20 pm
List:org.oasis-open.lists.wsbpel

Danny,

Sheesh. I'd missed all this going on. Suddenly the fog clears - and I see lights!

Can I suggest that you check out the CAM approach here? It's based on the Exampltron approach. So rather than inventing all this structure manipulation stuff - you just laydown the varients - and the context rules and then its MUCH easier to do - as you operate directly on the structure instance you want.

See the two parts of CAM

<AssemblyStructure>

and the

<BusinessContext>

for how we approach this.

Seems to me you could just put the structure layout in the BPEL and use the CAM predicates in-line - (just as Exampltron does) to instruct what is needed - prune, insert, etc.

What's nice about this is that you have a subset of the CAM spec's and therefore no need to reinvent all that - and then if people want more later on - you've already got the core means to do all this structure manipulation based on the content assembly rules that CAM uses (also XPath based). As we are all swimming in the same direction - based on established Exampltron / XPath technicques.

Does not make much sense to have two mechanisms for the same thing here.

Apologies for not "getting it" and making the connection here much earlier than just tonight in the small hours!

Cheers, DW. ================================================================ Message text written by INTERNET:chri@active-endpoints.com

Using the insertBefore operation.

The description from your proposal:

<insertBefore> <from/> <to select="XPath Expression evaluates to one node"/> </insertBefore>

creates a sibling before the select expression

Example 1:

<insertBefore> <from expression="'hello'"/> <to select="/foo/blah"> </insertBefore> <