Tony wistfully writes:

> Maybe if we ever get back to doing front-end compiler work for the HP3000 we
>  could explore this again? I know it's not standard, but WHAT a competitive
>  advantage it could be.....

Let me say that although I have never programmed a single line of COBOL in my
life (and hope to finish it out in that state), I could not agree more
strongly with Tony's comments and conclusions.

Standards aren't all they're cracked up to be. They may as well be called
lowest common denominators. They "level the playing field" by bringing a
uniform mediocrity to whatever they standardize.

I personally prefer to program in BASIC as an upper level control/sequencing
structure and a macroassembler as a means to generate high-efficiency code. I
have come to really enjoy this particular approach. Sequencing code (the stuff
where the behavior of the program resides) can be put together at extreme
speeds with very high reliabilities and yet the nuts-and-bolts macroassembler
code that does all of the work (and where the program actually spends most of
its time) can be made to be quite optimal.

A second attribute of BASIC is that no one has ever pretended that there's a
standard. Over the years I've used Altair BASIC, Rocky Mountain BASIC, HP3000
BASIC, Microsoft BASIC, and Visual BASIC. Some of these languages are
different enough that the might legitimately require new names. Some are
tokenized, some are interpreted, and some are fully compiled BASICs. But each
has their strong points -- and those strong points are generally aimed at the
system on which they're resident. Learning those strong points on each new
system generally promotes extremely efficient code that is exceedingly easy to
put together. If Tony's suggestion for COBOL were implemented, it would no
longer be standard, but the question is: "who cares?", especially if the
choice is being forced on to some "truly standard" version such as MicroFocus
COBOL that has no intrinsic understanding (pun intended) of the HP3000.

Again, I strongly endorse Tony's suggestion. Doing something well may be non-
standard -- but then again, standard is just another word for average.

Wirt Atmar