RE: [end points comm] OPES System

[jfcm <info@xxxxxxxx>]

At 17:48 20/08/03, Alex Rousskov wrote:
> On Tue, 19 Aug 2003, jfcm wrote:
> Things that are clear to most people in the real world become very fuzzy 
> in the virtual reality that we are building and documenting here. Not to 
> mention that I disagree with your assumption that
> ownership and domain of responsibility are always clear in real world.

Sorry, these are comments. Not real logical facts.

We have a hierarchical identification:
- top is system
- below is domain
- below is server
- below is service
- below or aside is the traveler

You may certainly imagine identification plan/protocol violations. But 
please come document them.
In the model we chose (airtransport) as seen from the users (not from a 
single user _and_ from a tour operator):

- the system (alliance, tour, travel) is identfied by the Warsaw 
conventions and the issuer of the ticket
- the domain (airline) is idenfied by its international code
- the server (plane) is international identified by its indicative and IFF 
or eventually destroyed (what you expect from an hacker)
- the stewart, the offered bottle, the milleage plan, the luggages, the 
passengers, are idenified and traced in many ways, sometimes there are 
confusions but that works usually well because protocols are pretty clear.


OK I accept that when we started I asked if OPES were ONES or not and I 
gave air reservation and  SITA attached systems as a ONES example. Or Swift 
related systems.

So I am ready to accept that OPES do not qualify to support that type of 
need, if you say so. But I do not see why.
jfc












> The "take them to court and win the case" answer is not satisfactory
> to me. I want the specs to be clear how the responsibility is
> assigned, traced, and managed.

This was not a response to a technical problem. Only a comment on the way 
the question was meaningless because you confused two layers.

You cannot demand a protocol to respond a question it is not to be asked.