On Monday, Sven, Sebastian, Moritz, Jan and me traveled to Cambridge where the Code Generation Conference takes place each year. Arriving one day ahead we had enough time to acclimate with fish ‘n chips and some ale at the local pubs before polishing our presentations.
The first conference day on Tuesday…
…started with a smooth introduction on modeling, DSLs and all the other terms specific to our focal point. It’s always interesting to see how other experts try to explain what we are doing and Juha-Pekka Tolvanen (MetaCase) did a great job even though he nearly ignored alternatives to graphical models.
Second, Sven and Sebastian talked about challenges in textual DSL design and concentrated on topics such as reusing languages and using models beyond the description of structures. If you follow their blogs you will certainly find out what they are interested in. This was the second time I noticed that the audience here at #CG2009 acts a bit more open-minded towards UML models and my coworkers sometimes had to play hardball with questioners.
Karsten and me skipped the next time slot to prepare ourselves for our own talk “Mastering Differentiated MDSD Requirements at Deutsche Boerse AG” we gave at the end of the day. We packed a lot material into 75 minutes but still managed to come up with a short demo that shows how Xtext can interact with graphical models. You can download the slides at slideshare but the demo is not yet ready, sorry (the attendees asked for this, please drop me a mail or comment on this post so I can inform you later on).
After a stressful day, the organized punting trip along the Cam made this day perfect to me.
…was a completely different story to me. Instead of listening to or presenting well-known ideas I had the opportunity to get some deep insights of concepts I only read about before. After an amusing staged fight between Markus Völter (textual modeling) and Steven Kelly (graphical modeling) my day started with the well-prepared hands-on session about MetaEdit+ Juha-Pekka Tolvanen, Risto Pohjonen and Steven Kelly gave. From what I have seen I would say that this software is the most sophisticated graphical modeling tool at the market today.
The session I have joined afterwards was not that impressive but I compensated for this with today’s highlight: Language Definition, Extension and Composition with MPS. Markus Völter and Konstantin Solomatov complemented one another and impressed the audience deeply with the upcoming version of MPS. This was the first time I have seen a live presentation on this little pearl of modeling.
I want to thank Mark Dalgarno and Andy Moorly for organizing this event. Questions about the room plan, the need for a different stage setup, giving up their seat at the dinner or ad hoc money exchange for thirsty attendees – no problems for them. Your upfront organization as well as your courteous appearance is simply overwhelming!
Update: This post has been published at The Model Driven Software Network, too.
- Slides of the session at slideshare
- Pictures of the conference at flickr
- Twitter tweeds about the conference