Steven Kelly on DSM

Code Generation 2007 was a rousing success, more than living up to my high expectations. The sheer number of top people from the DSM world was astounding: Steve Cook and Alan Cameron Wills of Microsoft, Markus Voelter, Krzysztof Czarnecki, Magnus Christerson, Xactium's Tony Clark and Andy Evans, Jos Warmer, Laurie Tratt, the usual MetaCase suspects, and Andrew Watson of the OMG -- plus many more from the wider MDD world. As Peter Bell put it on Twitter: "A well placed bomb at our table would have put back Domain Specific Modeling about 5 years -- amazing group!"

It was good to have Andrew there waving the UML flag: besides him it seems that most OMG people aren't keen on joining panels to discuss these issues. That's a shame, as I think it's important to have all points of view represented, even if some are unpopular with the audience. We were always happy to be involved in these discussions, even back when everyone would shout "bah, humbug, non-standard!" and pelt us with rotten tomatoes (not that we got many panel invitations back then, presumably because of the cleaning bills).

Maybe it's easier to take unpopularity if you are an upstart challenger, than if you've been the king of the hill and the world is now moving away from you. Or maybe it's easier if you're right, or at least think you are. It was interesting talking with Andrew over lunch to hear him say he didn't care about what was right, just what was "relevant". A little probing revealed that by "relevant" he meant "popular". That's a perfectly valid point of view, and one that will probably make you more money or even mindshare. But to me that's more like politics or entertainment, rather than science or engineering.

Here's a picture from the last session, where there were 7 flip charts to write the top questions and statements for various areas. Seated on the front row: with his back to the camera is Markus, then Jos, Magnus, and our own James Hammond. At one point James was holding forth, and everyone was hanging on his every word -- I wish I'd had a camera ready then! The interesting thing was the topic of the group, "Obstacles to Adoption". Few of the obstacles written on the chart were technical: actually getting your organization to start with DSM is often more of a marketing problem. And for marketing expertise, James is your man! You can know as much as you want about DSM and programming, but unless you also understand something about people and organizational change, introducing DSM -- or indeed anything new -- may well be an uphill struggle.