Accordingly to Douglas Purdy's post: From “Oslo” to SQL Server Modeling: OSLO will change as a chamaleon to be a different thing as originally promised.
I saw your tweet on this. It's too bad, but not very different from what happened with DSL tools: where others were trying to define DSLs for narrow, specific domains, Microsoft aimed for the ambiguous, broad domain of '.NET enterprise applications'. This is almost the same thing all over. Vlad could be right: there's no quick money in MDD for Microsoft right now, so why should they bother? I agree only partly on the education level, but bottom line I guess Microsoft sees no market pull for MDD, while it does see a market pull for modeling of databases. That's doing business...
I have followed Oslo since its very origin, because it seemed to be a drastic change and almost the first approach to a model driven point of view for designing from Microsoft; a more consistent way than Software Factories was, a return to UML consideration, and other known lines of development. But as soon it started its way, Oslo became oversold: too much promises, but few real stuff. I´m astonished on how a product with no clear goals and fundaments have created expectations, having adopters without almost anything to test.
Microsoft seems to be a giant without a head moving from side to side. I don´t agree that an erratic plan could be recovered easily. It would hurt its customer base, and shows a bad business model. Is it possible to commit with a tool that is abandoned a year later?
More than the rebranding issue by itself, I guess what really annoys JJ and a some of us in more or less degree is the over-hype generated during a year and half to finally change the course in the last moment and pretend it is the logical path to follow.
"The only real problem I see is that DSL and MDD are still niche markets and could never emerge !!"
I guess it all comes down to tools. Would UML be so widespread if there weren't tools like EclipseUML? ;-)
Projects like Xtext and MPS are also taking DSL's out of their niche.
So we need better and easier DSL and MDD tools if we want these approaches to be accessible to the wide developer audience. This is the first hurdle. The second one is to put the average developer in a "modeling state".
My analogy is related to the fact that big players are slow to react to changes. And software engineering IMHO is going through a(n exciting) period of change. New small but noticeable players will emerge as new technologies and methodologies evolve and be adopted by the industry.
What's wrong with my analogy is that most "dinosaurs" of our time have deep pockets, so they'll end up buying their survival. Just another fact of (business) life.