[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [ebxml-bp] WSDL / BPSS proposal
I favor the "Garden of Eden rewrites" (Venetian blind) which means we have both complexTypes and elements for them globally available. That also means substitution is easily available for elements as we evolve BPSS transaction type. Rather than inhibiting or disallowing the available extensibility mechanisms, I think we should support them. We can add any elements and anyAttributes if we need to if it does not water things down too much and if the grammars remain deterministic. Keep the X in xml! I also favor the explicit subclassing Martin proposes for BT. My only concern is whether we have some limitations in commonly used tools. However, lately I have seen much better performance of tools when substitution groups are used. The subtype constraint is being handled much better now than it used to be, and usually there are workarounds so that the "picky" implementations can be made happy. Dale Moberg -----Original Message----- From: David RR Webber [mailto:david@drrw.info] Sent: Thursday, February 12, 2004 7:33 AM To: martin.me.roberts@bt.com; anderst@toolsmiths.se; jeanjadu@Attachmate.com Cc: ebxml-bp@lists.oasis-open.org Subject: Re: [ebxml-bp] WSDL / BPSS proposal Martin, I like the concepts - but all this Venitian Blinds and abstract class overloading sounds too much like unnatural acts!?! Why don't we just support the 6 transaction patterns directly with a simple transactionType enumerated choice approach, and have set of associated criteria then selected and configured by that. Notice - if people then need subtle variations away from these 6 for local considerations - they can use the context mechanism to define those and agree them with their business partners. Since these extensions are agreed at the business level - this should go a long way to solving Anders concerns over the legal liability issues.. Thanks, DW. ----- Original Message ----- From: <martin.me.roberts@bt.com> To: <anderst@toolsmiths.se>; <jeanjadu@Attachmate.com> Cc: <ebxml-bp@lists.oasis-open.org> Sent: Thursday, February 12, 2004 6:23 AM Subject: RE: [ebxml-bp] WSDL / BPSS proposal IN my simplistic brain I am unable to carry all the variations that are being proposed for the various links to various possible standards. I therefore would like to make a simple proposal that might enable us to focus on the BPSS yet allow for sensible extensions. 1) the BPSS does not attempt to go beyond the process layer that we currently understand 2) We rebuild the schema to allow for hooks for extensions at most levels. This would include both the ability to replace an item using the Ventian Blind substitution group method as well as defined hooks for lower level artifacts, such as the WSDL definitions. 3) We produce a white paper showing how extensions work and possibly prime it with an extension for WSDL 4) we explicity define BusinessTransaction as an abstract class to be overloaded with subclasses for each of the 6 known transcation patterns, where the main difference is the default values for the flags - This would allow for two things a) overriding of flag values even when using a pattern and b) external new patterns for cases where the 6 existing do not work. This would be a major departure from 1.01 and 1.05 and even UN 1.1 but would lay a great foundation for other work in the future. What do people think? Martin Roberts xml designer, BT Exact e-mail: martin.me.roberts@bt.com tel: +44(0) 1473 609785 clickdial fax: +44(0) 1473 609834 Intranet Site :http://twiki.btlabs.bt.co.uk/twiki -----Original Message----- From: Anders W. Tell [mailto:anderst@toolsmiths.se] Sent: 12 February 2004 10:58 To: Jean-Jacques Dubray Cc: ebxml-bp@lists.oasis-open.org Subject: Re: [ebxml-bp] WSDL / BPSS proposal Jean-Jacques Dubray wrote: >You write that a OperationActivity is different from BT/BTA, since it >looks very similar to RequestResponseTransaction could you elaborate on >the reasons why ? >JJ>I responded to that in an ealier email (direction cannot be inverted >like a BT in two different BTAs with opposite roles) > >secondly how does OperationActivity handle semantics and obligations >related to dispatch, reach? > >JJ>What do mean with dispatch and reach? > > As defined by UNCITRAL Model Law on Electronic Commerce with Guide to Enactment (1996), with additional article 5 /bis/ as adopted in 1998 Article 15 and others. <http://www.uncitral.org/english/texts/electcom/ml-ecomm.htm> Basically these legal conventions and other electronic act points out the importance of separating sending and receiving of data messages. The "legal effect" may be defined at 6 point 1 Request.dispatch 2 Request.reach 3 Responce.dispatch 4 Responce.reach 5-6 Who has the risk when a data message has been dispatched and when it has reached is also relevant. All above considerations affects a BT outcome. It appears that specifying protocol error or method invocation exeception is not enough in an eCommerce environment. /Anders /anders -- ///////////////////////////////////// / Business Collaboration Toolsmiths / / website: <www.toolsmiths.se> / / email: <anderst@toolsmiths.se> / / phone: +46 8 545 885 87 / / mobile: +46 70 546 66 03 / /////////////////////////////////////
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]