Reply to Ian Joyner, regarding the futility of discussions about
computer languages.
I got a number of nasty flames by e-mail, so I'll respond to some
misunderstandings, and then leave discussions about languages to
folks with asbestos suits.
Yes, hard engineering discussions are indispensable to any
serious discussion of languages, and I have indeed read your paper on
C++. In fact, in its second edition, it was one of the things that
convinced me that there really _was_ something wrong with C++.
It is tough when you are an end user of a language. Since I'm not
a computer scientist or language professional, it is hard to state
_exactly_ why FORTRAN IV was harder to use than Pascal say, or why
C++ seems so much harder to me than ML. Nevertheless, I try hard
to stay out of language discussions, preferring to use the tools that
work for me and silently get the job done.
I also cheerfully admit that self-selected testimonials like the one
I posted earlier have no statistical validity. Any more than the
"everyone's using C++, so it must be the best" has validity.
The C++ people would do well to stay away from the bandwagon arguments
and the argument that says (or implies), "C++ isn't difficult; if you
can't master it at once, you must be stupid."
Of course, C++ is not all bad either. It did accomplish the goal of
adding object orientation to C. And low-level languages do have their
place. Sometimes you have to do some numerical task in an inner loop.
I also didn't mean to implicitly flame Ada. At least Ada was designed
consciously for a purpose. If you're going to write code for controlling
cruise missles, then safety ought be a very high priority indeed. I
admire the appreciation for human factor realities the Ada people always
seem to express, and I'd use Ada if I ever had to use a procedural
language.
OK, enough.
--Travis Porco