Re: asn06/OWL vs. UML [was: asn06 take 2 (Abstract Syntax as a kind of ontology?)]

Gerd Wagner wrote:
>>>> [By the way, how does "list of X" work in UML?]
>>> I don't see for what this would be needed in an abstract syntax?
>> In the abstract syntax we do need to identify places where 
>> ordering is significant. 
> 
> This is expressible in UML with the help of the {ordered} constraint, 
> see e.g. the REWERSE R2ML metamodel diagram
> http://oxygen.informatik.tu-cottbus.de/R2ML/0.4/metamodel/R2MLv0.4_files/png
> _23.htm

Makes sense, thanks.

Where is that defined? I tried to look for it in the UML 2.0 
superstructure specification [1] but the only entry for "ordered" (at 
least in the index) was to do with queue ordering, and I couldn't 
finding it in UML 2.0 OCL [2].

My ancient copy of Folwer [3] claims there is no defined syntax for 
these constraints. He recommends terms like "ordered" "bag" "ordered 
bag" and so forth but, at that time of writing, those appear to be just 
his conventions. Is there some standardized set of such constraint terms 
these days?

Sorry to be still showing my ignorance here, but if we are going to 
adopt UML as a formal notation, rather than just a diagramming 
convention, then I'll have to find some way to climb further up the 
learning curve.

Dave

[1] http://www.omg.org/docs/formal/05-07-04.pdf

[2] http://www.omg.org/docs/ptc/03-10-14.pdf

[3] UML Distilled. Martin Fowler, Kendall Scott. Addison Wesley 1997

Received on Wednesday, 22 November 2006 12:22:12 UTC