Some time ago I talked to you about our work on creating a profile mechanism for general EMF models (see “from UML profiles to EMF profiles and beyond” , including the full paper and the tool support).
We have now presented this work at the TOOLS conference. If you are interested and want to take a quick look at what we propose you can see now the slides that summarize the work:
FNR Pearl Chair. Head of the Software Engineering RDI Unit at LIST. Affiliate Professor at University of Luxembourg. More about me.
Got a nice comment for this work: “Found “From UML Profiles to EMF Profiles and Beyond” practically very ‘useful’, since it provided a good point in the DSL vs Profile debate, showed some interesting techniques for working on
the meta level, and it provided a tools that leverages an existing
platform.”
Hi Jordi,
I was wondering: do you think the Metamodel Extension Facility (MEF) RFP that is under discussion at OMG (proposed by Steve Cook, Microsoft) can have some kind of impact on the work you presented here?
(I’m mentioning it in my writeup of the last OMG meeting: http://www.modeldrivenstar.org/2011/06/some-highlights-from-salt-lake-city-omg.html)
Marco
Marco, this is a very good point. I haven’t read it yet but you are right, probably our work fit somehow there.
Nice work!
The re-use aspect is interesting but at the same the only thing that really adds value, as otherwise you can always use “downstream” (from your DSML(s)) languages which annotate instances of those in a concrete syntax. Especially with a modern LWB, this is fairly easy to do.
+ I’m still afraid that we’d be seeing the same thing as for UML: Profiles becoming the actual meta model, instead of the “real” Ecore model.