Xref: utzoo comp.edu:3726 uw.general:1966
Path: utzoo!attcan!uunet!ogicse!milton!iho
From: iho@cac.washington.edu (Il Oh)
Newsgroups: comp.edu,uw.general
Subject: Re: Recursion Summary
Message-ID: <10019@milton.u.washington.edu>
Date: 26 Oct 90 20:28:00 GMT
References: <1990Oct23.211651.10227@contact.uucp> <9868@milton.u.washington.edu> <9882@milton.u.washington.edu> <9893@milton.u.washington.edu> <9898@milton.u.washington.edu> <9917@milton.u.washington.edu> <90Oct26.142612edt.6786@neat.cs.toronto.edu>
Sender: news@milton.u.washington.edu
Reply-To: iho@akbar.UUCP (Il Oh)
Distribution: na
Organization: University of Washington, Seattle
Lines: 29

In article <90Oct26.142612edt.6786@neat.cs.toronto.edu> mdhutton@theory.toronto.edu (Mike Hutton) writes:
>
>In article <9917@milton.u.washington.edu> iho@akbar.UUCP (Il Oh) writes:
>>If you accept my claim that there's no recursion at the fundamental
>>logic level, you must agree that, theoretically, a recursive solution
>>can never be more efficient than an iterative one.
>
>Why does efficiency always seem to imply run-time cost?  The 
>dominant cost in most (educated guess) applications is the programmer
>time, not computer time---not everybody is doing combinatorial problems
>or number crunching.

Touche!  I seem to have adopted the narrow-minded assumption that CPU
time is more valuable than my own time.  I now see the error of my ways
and will charge more for my own time accordingly.  :)

>You seem to imply that recursive programs are a tradeoff of elegance
>for (time) efficiency.  I would suggest instead they are a tradeoff
>of programmer time and effort, maintenance and correctness for a 
>small constant factor of time which is probably not needed anyway.

I like your way of looking at things better.  However, the central
point of argument is still the same.  You can (though you may not
want to) do anything iteratively that you can do recursively.
--
  "Gosh!  You've really got          |    Il Hwan Oh
     some nice toys in here."        |    University of Washington, Tacoma
         -- Roy Batty, Bladerunner   |    iho@cac.washington.edu
                                     |