FW: Issues with the MetaGME paradigm

[GME]

----------
From: 	Tamas Paka
Sent: 	Tuesday, March 25, 2003 2:39:17 PM
To: 	GME Group
Subject: 	RE: Issues with the MetaGME paradigm
Auto forwarded by a Rule

Well, yes...
If you have to supply a kind, it is recommended that you don't type it as string (because those calls are obsolete).
In that way, mis-spelling can be filtered immediately in the time of writing expressions as well.

Of course... when you must supply a rolename, you shall continue to use strings.
Unfortunately, "connectionParts" supplied by rolename won't work, use "parts" instead.

Tamas

> ----------
> From: 	Jonathan Sprinkle
> Sent: 	Tuesday, March 25, 2003 2:18:06 PM
> To: 	GME Group
> Subject: 	RE: Issues with the MetaGME paradigm
> Auto forwarded by a Rule
> 
Thanks Tamas!  Should I do this for all parameters that I used to pass as a string now?  (e.g., models( "modelname" )) ??

> -----Original Message-----
> From: GME Group 
> Sent: Tuesday, March 25, 2003 2:15 PM
> To: GME List
> Subject: FW: Issues with the MetaGME paradigm
> 
> 
> 
> 
> > ----------
> > From: 	Tamas Paka
> > Sent: 	Tuesday, March 25, 2003 2:14:37 PM
> > To: 	GME Group
> > Subject: 	RE: Issues with the MetaGME paradigm
> > Auto forwarded by a Rule
> > 
> Jon!
> 
> The bug is fixed...
> 
> Until the next release of GME... 
> DO NOT use the call : gme::Model::connectionParts( kindName : 
> ocl::string )
> REPLACE with the call : gme::Model::connectionParts( kind : 
> ocl::Type )
> 
> so... 
> for example:
> 	the problematic call is:
> 		connectionParts( "DefaultTransition" )
> 	the good one is:
> 		connectionParts( DefaultTransition ) 
> 
> in specific:
> 	you have to do find the constraint called 
> "MaxDefault_1", which is attached to the model "Test" in 
> "TransformationLayout" sheet
> 
> Tamas
> 
> 
> -----Original Message-----
> From: GME Group 
> Sent: Tuesday, March 25, 2003 1:48 PM
> To: GME List
> Subject: FW: Issues with the MetaGME paradigm
> 
> 
> 
> 
> > ----------
> > From: 	Tamas Paka
> > Sent: 	Tuesday, March 25, 2003 1:47:41 PM
> > To: 	GME Group
> > Subject: 	RE: Issues with the MetaGME paradigm
> > Auto forwarded by a Rule
> > 
> I found it.... there is something wrong in the ConstraintManager..
> I am on it...
> 
> sorry, for the incovenience..
> Tamas
> 
> -----Original Message-----
> From: GME Group 
> Sent: Tuesday, March 25, 2003 1:26 PM
> To: GME List
> Subject: FW: Issues with the MetaGME paradigm
> 
> 
> 
> 
> > ----------
> > From: 	Jonathan Sprinkle
> > Sent: 	Tuesday, March 25, 2003 1:26:19 PM
> > To: 	GME Group
> > Subject: 	Issues with the MetaGME paradigm
> > Auto forwarded by a Rule
> > 
> I have a fringe paradigm that I cannot get to work under 
> GME3, with the following "fringe" components:
> 
> 1) Libraries
> 2) Libraries of metamodels (that of course, now that we have 
> MetaGME instead of MetaGME2000 have changed)
> 3) Multiple Inheritance
> 4) Inheritance by Proxy
> 5) Aspect inclusion by role (rather than kind)
> 6) Minnie Others
> 
> I have included the paradigm here (although I don't know if 
> includes are allowed on this list) for anyone who wants to 
> play with it.  My question is this: can anyone who has GME in 
> debug mode tell me why when I try to create a new GME project 
> using the interpreted paradigm that i immediately get an 
> emergency-event-thanks-for-playing message??
> 
> Jon
>