This piece raised some interesting questions for me.

I’m going to try and avoid the obvious response which, as I work for a development tools vendor, you can probably guess.

But I’m wondering at what point do you decide that your tool is doing quite enough thank you very much? I know this will be different for everybody, but why is today’s level of abstraction enough, whereas the level we had 5 years ago was not? Why was it OK for dbExpress to abstract out details of dealing with databases, but it’s not OK for ECO to take that further?

Is the rise of code-focused features (like refactoring, smarter code templates, code intentions, etc) in IDE’s a recognition that giving people more wizards, components and frameworks is not necessarily what they are after? Just give me stuff that lets me bang out code faster?

I was just chatting with someone about this who said that your comfort level is determined by the technologies at which you first become truly proficient. Actually, he qualified this by saying this was the point where you no longer had to spend too many cycles on the details of the framework, language syntax, etc and could focus the majority of your attention on the problem. For example, I used to be a C++ developer, but I really sucked. I jumped ship to Delphi 1, but I don’t think I approached this "proficient" point until around the Delphi 5-ish timeframe. So, this theory says that Delphi 5 should be my comfort level and everything after that just becomes "this bloody tool getting in the way".

Hmm, don’t think I buy that. I’m not sure what I would buy instead, but this doesn’t do it for me. Or is it that you reach a point where it’s just hard work to keep up with the latest doo-hickey and you either put on a suit and tie or grumble about Microsoft taking the fun out of programming.

Well, I’m wearing a suit and tie as I write this, so I’m sure that’s not it 🙂

However, it is an interesting thing for tools vendors to be thinking about. I know I’ve faced this type of response from people while we’ve been out evangelising .NET via C#Builder and Delphi 8. Our industry is pretty much built on raising productivity by lifting the level of abstraction, and I’m not suggesting that should stop. Is it just a case that, as Geoffrey Moore would suggest, eventually market pressure will force the the laggards to adopt the new abstraction or get left behind?

Theories anyone?