The Model Driven Software Network

Raise your level of abstraction

What are the main trends in MDSD - according to you?

Looking over the past few months, say since PPL 2009, I perceived the following trends in model driven development.

  • language workbench is a term that seems to be replacing the notions of MDA-tool and DSL-toolkit very rapidly now. Of course the term is not new, but it is becoming more common.
  • in DSL world, the distinction between graphical and textual DSLs is rapidly being replaced by integration of the two, see cooperation between Itemis and MetaCase, but also recent attempts to integrate XText with GMF
  • language workbenches with multiple views on the same content, like Intentional has been promoting, are showing up in multiple places right now (MPS, Spoofax, the integration mentioned above)
  • given what I read in varoius articlesin  the ACM Journala couple of months ago IEEE Computer ed. October 2009, models@runtime will become more visible shortly as well.

Which trends do you, MDSNetwork members, perceive, both in practise and in research? I'd be glad to hear from you, since I've been asked to do an article on trends in this area (for a Dutch magazine) over the next few weeks. Of course an English translation will follow shortly afterwards on my blog.
I'd be more than happy to receive ideas here, in mail, through twitter or face-to-face during CG2010 next week.

Views: 186

Reply to This

Replies to This Discussion

Nice observation, and probably true in a lot of organisations. Would you care to elaborate a bit?
If he elaborates his answer, he will say that's because of all these crappy open source products. Right?
I'm shocked by your scepticism ;-)
So was I, but I'm still waiting for Vlad to respond ;-)
It's fully wrong.

My company (Obeo) has a commiter (Mariot Chauvin) on GMF and we are still active on this project, with some IBM guys and some external contributor which provide lots of patches.

GMF is very dynamic and now, SAP join us with a new set of API for the backend (project named Graphiti).
And GMF is of course fully compatible with Helios.

And you also are wring about Eclipse Modeling: it's work, it's broadly used, and it's stable (more stable than other old tools like Omondo, for example).

And take a look at Acceleo and Xtext, some new project in Eclipse Modeling, to see how active is the community and how we success to do exemplary tools. And as I know, Itemis, Obeo and their partners have no financial problems. And you? :)
Thanks Etienne. Do you have any idea where the rumor that GMF has no committers comes from? Vlad mentioned it, but there are others as well...
Yes, I have an idea as my company pay a Mariot to be a GMF commiter ;) (he is dedicated to performance issues)

Today, RIM, IBM, Redhat, Obeo and some individuals are commiters.

You can see contribution on the Eclipse dash :

And :

We have launched of product to create easily some Domain Specific Modeling (Obeo Designer, and it's based on GMF Runtime and we success to manage very large and complex models.
Vlad doesn't like OpenSource, because he doesn't success to find a business model. Our is simple: you can develop with GMF by yourself or with training and expertise from Obeo, or if it's too complex or too expensive, you can use Obeo Designer which come with a kind of "configurator".
Some suggestions of things I see:

- Model execution is a viable alternative to code generation. Depending on your target users it can have a lot of advantages.
- Cloud computing will be a game changer. A lot of cloud platforms have MDD like tools to build or configure applications (they can learn a lot from existing MDD tools about DSLs and modeling techniques). See also my recent article on this subject:

Missing trends in my opinion (i.e. we need more of this):
- Focus on business results instead of technology.
- Focus on methodology (how to use the tool) and changes in project management (not everyone - understatement - has experience with 'agile').
- ... curious to hear your ideas! ....
Nice observations, Johan. Indeed a lot of things are about technology rather than business and methodology, Tony Sloos, Richard van der Laan and myself briefly touched on that in at the Dutch Agile MDD seminar in March:

Also, model execution is one of the drivers behind models@runtime as well, and I'm working on a possible product in this area as well. I'll have a look at your cloud computing article as well.
No problem, this is a forum after all, not a chat box. And, actually I was enjoying a bottle of rosé from the Languedoc when I wrote my reply.

Anyway, I think you are right in saying that there is a market focusing on web services and UIs, which may not benefit too much (nor be interested in) MDSD due to it's current focus.
However, there's also a number of other markets where things may be different. Looking at the automotive solutions, and activity in some high tech industry companies that I'm aware of I think trends in MDSD (or away from them) may vary per market. Hence the generic question in my initial post, and many thanks for sharing your vision on the market of big corporations and the web services domain.
There is 2 types of M2M transformations:
- a model (that conforms to a first metamodel) is translated into another model (that conforms to a second metamodel).
- a single model is modified directly.

From now, we focused much more on the first alternative. The second one opens some possibilities such as :
- models@runtime
- rule based transformations

For the models@runtime, everybody tends to agree, it will be one of the next MDE buzzword.

In the domain of rule based transformations, I'm more and more convinced of graph based transformations, these transformations have a powerful expressiveness.

Another trend could be live & declarative UI, that seems to grow up slowly but surely with projects like XWT/E4 or wazaabi.

(In fact I'm dreaming about an app with a business model and a UI model weaved at runtime by graph based transformations !)

I also think MDE should be a convenient way to describe a process and automatically adapt your IDE to your methodology (something like Jazz/Rational Team Concert technology).

Finally I'm a bit disappointed that we don't see much modernization techniques (maybe I don't know them, so let me now!) such as MIA-Modernization.

I think we are at a point where we realize that MDE is not only about code generation, and MDE will amaze us!


Xavier Seignard
Thanks for sharing, Xavier. Looks like we share at least one dream...




© 2019   Created by Mark Dalgarno.   Powered by

Badges  |  Report an Issue  |  Terms of Service