Oslo and the two edged sword
Gar Mac Criosta has a series of posts on the SOA/BPM conference and one of them deals with the orchestration generation that was demo-ed. He makes a really good point about this being useful and dangerous at the same time.
As I commented on his post, IMO, this was always a problem with the ODBA. Business Analysts do not have enough tech knowledge to design orchestrations and we shouldnt be trying to flatter them into thinking they do. However, some biztalk architects use the ODBA for the design planning meetings to get consensus on the orchestration patterns, key paths through the process etc which makes more sense than having a BA use the tool. Thats why Oslo tools could be a double edged sword. Being able to generate orchestrations from models that we can quickly draw when working through designs with colleagues can be useful to get the solution structure in place. Turning that into production code is entirely a different matter and having a business person make that decision is to me, a definite no-no. Or, as that new dodgy song goes “No-No-No”.
What do the rest of you think? Have you used ODBA a lot? Have you managed to get the BA’s doing things at the right level of detail?