Some thoughts about GMF [message #205663] |
Wed, 17 September 2008 11:26 |
Eclipse User |
|
|
|
Originally posted by: mikko.laakso.avoidspam.com
Hello,
I've been working with EMF/oAW for just half a year now, and lately with
GMF and Metaedit+. Comparing the two, EMF-related tools lose to Metaedit+
on ease and flexibility of creating graphical editors, and wins on the
generator and standards compliancy (XSD etc), and metamodel editing. It's
still more important, for me at least, to get a good graphical editor
easily, than to create a flexible metamodel and generator easily and
suffer penalty on the notation side. This is one thing GMF is trying to
address and although there is a lot of progress, there is still much to do
before GMF can really be the number one productivity booster, which it
could be :).
There are some features which would really benefit the modelers:
- tighter integration between Ecore tree view and GMF graphical view.
* The tree view can be really useful in itself, but it is difficult
to
see object associations. GMF offers easy way to create and see
associations, but lacks the completeness, scalability and
versatility
of the tree view. So it should be possible to use them together
fluently. This would be an advantage for GMF.
* It is common that metamodel is quite complex, since it is usually
good
to take the complexity to the metamodel instead of the generators.
Because of this, and scalability issues, we simply dont' want to
(or
can't) use GMF to draw complete Ecore model, but part of it.
- Several views for one model
* we could have a single model file for versioning, but still be able
to
give different views for different users, also for dividing a model
to
keeping the notation simple enough for user. It should be a part of
GMF
functionality.
- Decomposition diagrams
* I think, scalability is an issue in many current models. With a lot
of
associations, they become unusable when the size of the model grows
very large. Why can't we just click an object in GMF-made editor, to
see it's containing diagram? This is an approach adopted by
Metaedit+, and I think it's brilliant for coping with scalability
problems.
--
Regards,
Mikko Laakso
|
|
|
|
|
|
Powered by
FUDForum. Page generated in 0.03690 seconds