Re: SOLVED! Decoupled Mutual Recursion Challenger

[bobduff@dsd.camb.inmet.com (Bob Duff)]

In article <38pulp$ovg@oahu.cs.ucla.edu>,
Jay Martin <jmartin@oahu.cs.ucla.edu> wrote:
> ...There currently not even one class
>being taught about Software Engineering at my university (not this year or
>last) Ada isn't even on our CS dept computer, C++ probably wouldn't be if
>it didn't automatically come with GCC.

I've heard that the GNAT Ada compiler will also be distributed with gcc.

> ...Look what I wrote in a boolean expression in C today:  
>BoolVar = (IntVar =! 2);.  I wanted  "!=" not "=!".

Interesting bug.

>If you design a language that requires 200 IQ brainos who never make
>errors to understand/use, then trying to use that language on a
>project of a hundred average programmers is not going to be
>successful.  Thus language design must be down to earth,

All programmers make mistakes.  Even those with 200 IQ.
I agree -- language design must be down to earth in the sense that
languages ought to prevent and/or detect down-to-earth mistakes.

> ...Of course, social science programming experiments could
>be performed to prove that certain language styles are more SE efficient
>than others but this might step on some toes and besides its obvious
>and boring. 

I think the reason such experiments are not done is that they would be
very expensive.  We can't experiment on rats, after all.

Sure, theoretically, you could compare Ada-with-discriminants to
Ada-without-discriminants, using two groups of 100 teams of programmers,
one for each language, all doing the same project.  But that experiment
would cost hundreds of millions of dollars.  And that's just to
investigate whether discriminants are a good idea.  What about all the
other features of various languages?

You could compare Ada with C++ using a similarly costly experiment, but
what would that tell you?  Having found out which of the two is "better"
a language designer would still want to know why.  Surely neither one is
perfect; we should strive to improve on both.

And, of course, nobody in "real" science trusts experimental results
until they've been duplicated by several different researchers.

That's why I think we're going to be stuck with seat-of-the-pants
language design and anecdotal-evidence on their merits for at least some
decades.

They say that in computer science, one takes a single data point, and
extrapolates from there.  It makes curve fitting easy.  ;-) ;-)

- Bob
-- 
Bob Duff                                bobduff@inmet.com
Oak Tree Software, Inc.
Ada 9X Mapping/Revision Team (Intermetrics, Inc.)