Productivity in software

As I will have to give two more seminars on software production in the engineering department, I would like to refine my previous notes on software productivity. Now, which are the principles guiding my development ?

Well, the first and actually the only one is quality of the result. This is strictly linked to careful and continuous modeling of the matter at hand. Tools here are a mixture of intuition, practical sense, experience, some formal knowledge, a lot of technical knowledge, and a good object oriented modeling capability. So I assume that modeling is sacred, and most of the energy should be towards this.

By quality I mean a complex notion, which is not quality just from the developer' point of view (which includes the designers'), but also quality from the systemists and customers points of view. And by reactions I don't just mean immediate reaction, but real and satisfied usage of the software for many years. Now from such high ideal may consequences follow, of which at the moment only a few come to mind:

- concentrate on the business model, which may generate an architecture, not on the architecture first
- know the best tools which relive you from syntacticalities
- never write the same concept formlization twice in code
- always use strong typing

Now from these general consideration technical choices follow too. [to be continued]