- From: bhaugen <linkage@interaccess.com>
- Date: Tue, 06 Aug 2002 12:47:42 -0500
- To: www-ws-arch@w3.org
Hal Lockhart wrote: > As I stated in a previous email, I agree with the general notion of avoiding > ambiguity, however I think Mark's idea of not depending on anything external > is unrealistic. Every business transaction depends on shared understandings > about the goods or services involved, the terms and conditions of sale and a > host of other things not explicitly spelled out in the network messages. Of > course, many of these are defined by centuries of commercial practice as > well as laws and regulations. However, others may vary depending on the > legal system, industry and over time. Ok, you're right. I took Mark's statement in its original context, which was default values popping in from schemas, that is, alteration of the message contents. If you take it in the context you explained very well above, it's not true. But for all of the other dependencies, if they are Web resources hyperlinked from the resource that the subject of the transaction (e.g. an order), then it would still be possible to have a definite answer to enough questions to be able to do business. > I believe the most practical approach, which is currently happening, is for > industry consortia to establish standards for the syntax and semantics of > common transactions in their industry. By making use of these standards, it > should be possible to avoid a semantic misunderstanding (deliberate or not). > A party who claims to have reason to use semantics which are contrary to the > established standards for the relevant industry will face a very difficult > burden of proof. Most but not all of the industry consortia are focusing on document contents alone. The business processes, transaction handling rules, etc. also need to be spelled out for anything resembling proof. For a simple example, did the seller agree to be bound by the buyer's order, or not? How can you tell from the order document alone? Is an explict acceptance response required? What if the response comes after the price is reduced? What if the buyer cancels the order? What if the seller delivers the goods later than promised? ...and on and on... -Bob Haugen
Received on Tuesday, 6 August 2002 15:58:56 UTC