[plt-scheme] what's HtDP, what it is NOT

2009/2/23 Henk Boom <henk at henk.ca>:
> 2009/2/23 Grant Rettke <grettke at acm.org>:
>> On Mon, Feb 23, 2009 at 8:03 PM, Matthias Felleisen
>> <matthias at ccs.neu.edu> wrote:
>>> Alexander also hates, absolutely hates, how computer scientists
>>> misappropriated his patterns. If you ever have a chance to read his original
>>> Notes, cover to cover, you will see a lot of similarity with HtDP and
>>> believe it or not, Robby's contract work. -- Matthias
>>
>> Which original notes? Are they in a book?
>
> I'm guessing that he's referring to "Notes on the Synthesis of Form".
> It's an excellent read.
>
> That being said, I do remember something I found distressing in it.
> For some inexplicable reason, it's not sitting on my shelf in front of
> me at the moment, so I can't double check.
> ...

Found it, it was wedged out of sight by two other books. the preface I
was talking about is called "Preface to the Paperback Edition", and
after re-reading it, I realize that I had previously understood even
less than I thought.

My newer (and hopefully more accurate) interpretation of the foreword:

What he encourages is not patterns ("diagrams") themselves, but the
idea that a problem can be split into many small parts which can be
solved individually.

What he discourages is the treatment of the design process as a purely
mathematical process.

As far as I can see, he says that as long as you can segment the
problem into pieces with few interactions between them, the process by
which you come up with solutions to the individual pieces is
relatively unimportant. I'm not sure I agree with him here, since
(especially in software) the individual pieces are not trivial to
implement, and can't always fit in the mind at once themselves.

    Henk