[gme-users] Problem with OCL

[Larry Howard]

Let me take an initial pass at this.

First, my understanding is that constraints are defined on the paradigm
as generated, not as modeled.  Abstract FCOs are part of the paradigm
definition (meta-model) and do not exist in the generated paradigm.
Therefore, it is not possible for a valid constraint to refer to them.
(Someone else should comment on constraint validation in the paradigm
generation process.)

That said, the result is less power in constraint definition, since the
constraints need to be defined on the concrete FCOs.  I must also leave
it to others to say what intentions there might be for meta-modeling
refinements that could address this situation, or to correct any
misunderstanding on my part reflected in this response.

-----Original Message-----
From: gme-users-bounces at list.isis.vanderbilt.edu
[mailto:gme-users-bounces at list.isis.vanderbilt.edu] On Behalf Of
Christian Brunette
Sent: Friday, March 18, 2005 2:55 AM
To: GME users
Subject: [gme-users] Problem with OCL


Hello,

I'm a new user of GME. I have a problem with OCL and I don't know if it 
is a bug or a feature not yet implemented or a constraint in OCL. I have
define several FCOs which inherit from an abstract FCO in which 
some attributes are defined. I add a connection whose source and 
destination are this abstract FCO. I would add a constraint on this 
connection. But I need to access to the attributes of the connected 
FCOs, so I need to cast them. But the constraint checker said me that I 
can't cast because the FCO is abstract. Is it a bug ??

Sincerely,
Christian

-- 
Christian Brunette
Post-doctorant in the ESPRESSO team
IRISA-INRIA, Campus de Beaulieu, 35042 Rennes cedex, France
Phone: +33 (0) 2 99 84 72 31, Fax: +33 (0) 2 99 84 71 71

_______________________________________________
gme-users mailing list
gme-users at list.isis.vanderbilt.edu
http://list.isis.vanderbilt.edu/mailman/listinfo/gme-users