Re: Scheme for programmers book

David Rush wrote:
> Michael Erdmann <michael.erdmann@snafu.de> writes:
> 
I have already some fun with it while trying to get schmoz working.
Bigallo does not understand the (require, DrScheme does but
complains about (define (function) ....) and Chicken is not able
to compile the SLIB library. Okay let's see, i have just started.

> ... 
>
> Have fun. You use a *lot* of parenthesis in Lisp. If you follow other
> language conventions you'll have a lot of lines with nothing but
> parens on them.
>
I am beaten, i will follow the convetions since everbody tells
me to do so .. :-)
> 
>
> This is also true in Scheme, although most compilers don't (Stalin is
> the only one that does)
.
It seems the compiler has vansihed from the net, for example:

http://www.neci.nj.nec.com/homepages/qobi/software.html

is not exisiting any more. They have completly reorganize everthing.
Any idea where i can find the source code?


> In order for you to understand this you're
> going to have some work to do first. You're going to have to learn to
> define types in terms of behavior, rather than data structure. Since
> this is the proper mathematical definition of a type anyway, you'll be
> better off for having done the work. Once you get to behavioral
> definitions, you'll find that Scheme has more wonderful tools built
> into the standard than exist in most other languages.
I think also, traning and actual understanding will be one critical
issue, specially when you have used classical main stream
programing languages for a long time.
I currently writing a small prototype of what i am intending to
implement in a more robust way later for common use. Lets see
how i will change my attitude to such of may major concerns.
 
> It's no worse than any other language I've used (except possibly SML,
> I've not worked in Ada) and better than many because of the functional
> definition of types.
My personal feeling is already it is even worse but let see :-/

>
> 
>
>
> It works surprisingly well, actually.
I am realy wondering when you are deploying a system to your customer
and something goes wrong, how is the application expected to behave.
Stop somewhere in the runtime system, call backtrace ...?

>.sigmonster has once again
> demonstrated his sentience. Mind you, most people fear freedom, and
> *all* managers do.
For a good reason, they are held responsible for the success
of a project, they can't simply point to same guy in the
team and say "..this guy feels now free as the wind, but we will
not be able to finish in time, so just pay the legal fees for the
delay ..". But this is a completly different storry. :-(


Thanks for all of the input. What i will do, i will play around
with fdscript (and scheme) to build my prototype.
Mean while i will collect all my results from this thread into
a small "paper" and put it for discussion in this group by
end of the w. Is it a good idea?

Regarding my fdscript decission i am currently working on it. I
think i have something important ....

Michael








>
> david rush