[comp.lang.icon] Icon as a teaching language

ron@mlfarm.UUCP (Ronald Florence) (12/23/90)

With the Christmas vacation upon us and a 12-year old around the
house, I wonder if anyone has had experience with Icon as a first
structured language.  My son is adept at vi and gnuplot, and for
homework problems I have occasionally kludged together awk scripts,
which he then modifies as he needs.  I suspect some of the more
elegant and terse Icon constructions could be daunting, but the
multiplicity of approaches to a given problem could be fun.

Has anyone tried this?  Any wisdom or suggestions?  
--

Ronald Florence			ron@mlfarm.com

wgg@CS.WASHINGTON.EDU (William Griswold) (12/27/90)

>From: mlfarm!ron@hsi.hsi.com (Ronald Florence)
>
>With the Christmas vacation upon us and a 12-year old around the
>house, I wonder if anyone has had experience with Icon as a first
>structured language....

Well, I was programming in SNOBOL at this age.  The results were fun
but aesthetically disasterous.  Certainly Icon would be at least as
fun and perhaps a better language to begin with.  If you can get him
playing with some of the more advanced functions, such as map and
random selection (kids eat this stuff up), he'll have a blast.
Actually, giving him something to modify could be very effective.
Small changes in small Icon programs can yield vastly different but
useful results.  The Icon books are full of small programs that are easy to
write and modify.


Don't forget numeric calculations as well.  Your child may be old enough
to program a sieve for primes (my first program divided by all the numbers
less than or equal to 1/2 the number.  It wasn't long before I figured out
I could leave out most of the even numbers....) or some other such program. 

Unfortunately Icon isn't interactively interpretive, so you can't
run-as-you-program.  I've written a little line-oriented interpreter loop
for Icon, but it changes the syntax and doesn't allow procedures....

					Bill Griswold

goer@ellis.uchicago.edu (Richard L. Goerwitz) (01/03/91)

In article <9012261840.AA07710@june.cs.washington.edu> Bill Griswold writes:
>
>>With the Christmas vacation upon us and a 12-year old around the
>>house, I wonder if anyone has had experience with Icon as a first
>>structured language....
...
>Unfortunately Icon isn't interactively interpretive, so you can't
>run-as-you-program.  I've written a little line-oriented interpreter loop
>for Icon, but it changes the syntax and doesn't allow procedures....

Has anyone written an Icon interpreter?  (Bill, could you post yours??)
It surely is the case that my son functions better using interpreters.
I also would not mind the ability to test out code fragments using an
Icon interpreter, even a rough and fairly fragmentary one.

-Richard (goer@sophist.uchicago.edu)

wgg@CS.WASHINGTON.EDU (01/04/91)

I'm afraid my interpreter is not for Icon, but a language much more limited
(and syntactically bogus) than Icon.  It was fun to build, but it shows
the limitations of line-based input in spades.

						bill