Senior members in our team have suggested that PM team should start using the DSL tools to prototype some of the new architecture models/designers that are being proposed. Consequently, I’ve been laboring away on my DSLs for the last few days and Pedro’s been a great help. However, there are two things that I (WARNING: Controversial statements about to ensue. Please read my disclaimer) do not agree with.
The first has to do with the target persona for the DSL tools. Which I was told was the Elvis/Einstein persona for at least v1. The belief is that you need to be at that level to be building DSLs. Well, I’m a Mort! As I write this, I wonder if I have just committed career suicide…:)
This, jokes aside, does not mean that I’m not smart guy. I understand my domain and I want to model the domain in order to get a designer up and running in to perform the tasks specific to my domain. I’m not interested in the DSL tools themselves per se, I’m interested in the solving my problem as quickly as possible. The idea is around RAD development for the prototype, so that we can incorporate feedaback as quickly as possible and get buy-in to spend the money to build the real tool. Which I leave that to dev team, with the personalities best described as Elvis and Einstein.
Pedro, after hearing this conversation in the hallway, forwarded me specs about the next interation of DSL tools. While I’m not sure that Pedro bought into the fact that DSL tools should be for the Mort user, the new upcoming features I feel are a leap towards helping Morts build their DSLs.