vestal (04/02/83)
I've just read some responses to an earlier article which evaluated
a few common programming languages. I would like to (flame-)throw
in a few comments in defense of those of us who are not omniscient:
1) My interpretation of the article was that a casual, subjective
evaluation was being made of the languages; no religious
percusion of the languages or their adherents was intended.
I personally found it somewhat more entertaining than
recursion vs. iteration, or parenthesis after return/sizeof, etc.
2) I don't think I need to be well-versed in the lambda calculus
to complain about matching parenthesis in LISP, or denotational
semantics to haggle about the best Pascal linear search loop.
I'll freely admit I've never written over 500 lines of LISP
code, and I personally don't think that IN PRACTICE the
language consists of only one data type and two flow control
constructs (or no data types or a whole lattice full of
data types or whatever you want.) This isn't a diatribe against
LISP, I just don't happen to think the formal underpinnings of
a language completely encompass everything relating to programming.
3) From a software engineering/ language design point of view
I think information like this can be useful. "Language
designers propose but programmers dispose", to make a poor
paraphrase. If we come up with a language that requires
programmers to know it's denotational semantics, God help
us.
Now in my experience, there's no subject for discussion which I've
seen raise tempers and ferment dogma like programming language
design. While I'm just as bad as most people at dismissing
what I know to be obviously crackpot ideas, I don't mind spending
a few minutes listening to/reading them (after all, it contributes
to my sense of superiority). I personally would like to keep
the tone of mortal combat friendly.