Saturday, June 20, 2009

More on authors' dilemmas

One of the authors' dilemmas about the right way to write ab event processing book is about - whether the book should concentrate around an "event processing language". One of the reviewers indeed suggested that the language will be centered about a specific executable language, and describe the concepts using the language, for the concepts that are not supported in the language, the reviewer proposed to include them in advanced topics section.

The idea is not new to us, when we discussed with the publisher about the book's structure, we considered such "bottom up" approach, however, we decided that it is methodically better to take "top down" approach. There are two reasons for it: the first -- we are building the book in a methodical way, and talking about the building blocks that are required to build applications, since different languages took different approaches and often the implementation blurs the concept, we thought the the bottom up approach will be a good way to teach a particular language but not the concept; the second -- the current languages follow various programming styles. We neither want to concentrate on a single programming style, nor to do a comparison, as this is not the goal of this book.

So - how are we approaching this issue:
  • We have constructed a descriptive meta-language that follows the concepts we are defining. We may also provide some visualization of this language (checking this issue now), and enable to play with this language on the future book's website.
  • We have asked all language owners (and some have responded positively) to provide their own implementations of the use case we are using in the book. We shall put code samples in all relevant chapters as provided to demonstrate the different attitudes. I'll write soon some comments on the use case, following some questions from language owners.

No comments: