RE: C Language

["James H H Lampert" <jamesl@xxxxxxxxxxx>]

> Anyone that doesn't call ["!"] a "bang" hasn't had e-mail long enough.
> Long time users will recall their email addresses as a "bang path". <G>

I'm surprised nobody's mentioned "Fencepost" for "|"

As to the C language, I blame its popularity on excessive specialization
among programmers: when the only tool you know how to use is a hammer,
then everything in the world looks like a nail. It seems to me that far
too many programmers choose to overspecialize, learning more and more
about less and less until they know everything there is to know about
absolutely nothing. Knowing everything about nothing may have qualified
one to write scripts for "Seinfeld" ("it's a show about nothing"), but it
doesn't make one a better programer.

Then there's the absurd notion (of which John Houston, my erstwhile boss
at Oxxi, was an exponent) that since modern GUIs supposedly reduce
programs to an event-driven string of one API call after another, the only
thing that should matter in a programming language is the ease of calling
APIs.

Then there are Niklaus Wirth's twin delusions: (1) that a programming
language should be as small as possible, and should compile in a single
pass, and (2) that it is both necessary for (1) to happen, and generally
desirable, to have mandatory separate operators for assignment and
equality-test. BOVINE SCAT! If a simple BASIC interpreter can tell whether
"X=1" is meant as an assignment or an equality test without benefit of the
obsolete "LET" reserved word (and all but the most primitive ones can),
then certainly ANY compiler can do so!

Then, of course, we have the completely backwards notion that development
boxes should be big and fast, while compilers should be small and fast.
This leads to disorganized "rabid" (not "rapid") prototyping, while
keeping programmers from even being AWARE that they're writing bloatware.

Thus we have C's popularity as a consequence of programmers not being
bothered to learn multiple languages (and therefore not knowing their
strengths and weaknesses), and being taught blatant academic fallacies
about programming language design, and not knowing when and where they
ought to drop down to assembler (if indeed, they bothered to learn any
assembler languages at all), so they settled on a language that's a
compromise between true HLLs and assembler, combining most of the worst
features of both.

And thus, we ended up with the two most popular object-oriented languages,
C++ and Java, being based on this, instead of on, say, PL/I.
--
JHHL