[gme-users] Doubts with MetaGME

Matthew J. Emerson mjemerson at isis.vanderbilt.edu
Fri Dec 16 09:12:12 CST 2005

I believe Kitty is asking for the ability to define more powerful
connection types which at the model level can compose other objects. It
would mean that instead of merely having <<Connection>> at the
meta-level, you would have <<AtomicConnection>> and
<<CompositeConnection>> or something.


-----Original Message-----
From: gme-users-bounces at list.isis.vanderbilt.edu
[mailto:gme-users-bounces at list.isis.vanderbilt.edu] On Behalf Of Akos
Sent: Thursday, December 15, 2005 11:14 PM
To: gme-users
Subject: Re: [gme-users] Doubts with MetaGME


I do not understand your question. Are you talking about MetaGME or GME?

In the MetaGME paradigm every object is an atom. There is no 
composition, no hierarchy. So, I am not sure why you would want 
especially a Connection (atom) to contain anything...

Or maybe you are talking about connections in any paradigm in GME not 
being able to contain stuff and having to work around this limitation 
during metamodeling?!

Pls. clarify. You may want to do it in person, my verbal communication 
skills maybe better than the written ones ;)


Krishnakumar B wrote:

>Hi Folks,
>Is there any reason that connections are (and should be) atoms in
>Quite often, I find that the elements that I use to describe a
>needs to be associated with properties, and in some cases needs to
>(i.e., composition) other elements.  And in almost every use of
>connections, I have to work around it by adding another element (either
>atom or a model), and making connections to and from that element.
>increases the number of elements as well as the number of connections.
>I would like to know if there is an elegant alternative to designing
>connections without the extra baggage.  I am also interested in any
>solutions that people have used in the past, i.e., something like
>Patterns" for metamodeling.

gme-users mailing list
gme-users at list.isis.vanderbilt.edu

More information about the gme-users mailing list