Help! Unemployed

Richard A. O'Keefe ok at atlas.otago.ac.nz
Tue Aug 14 00:30:04 UTC 2001


"Noel J. Bergman" <noel at devtech.com> wrote:
	One could just as easily suggest that anyone who spells it Fortran hasn't
	bothered keeping up with accepted use of the English language.

One could say that, but one would not be justified.
Someone's name is whatever they say it is.
The name of a programming language is whatever the "owners" of
the language say it is.
	
	ALGOL is an acronym.

The word "acronym" is ambiguous.  Some people use it exclusively for
words made up of the initial letters of other words, such as "NATO".
"ALGOL" not an acronym in that sense.  Some people also use it for
words made up of initial parts of other words, such as "Comintern",
from "Communist international".

The convention about putting the letters in upper case applies ONLY
to acronynms such as "NATO" that are made only of initial letters;
it does not apply to words such as "Comintern", which are not.

When Algol was the "International Algorithmic Language", it was
properly written IAL.  As just "ALGOrithmic Language", it may quite
properly be written "Algol".  Note that the
	REVISED REPORT ON THE ALGORITHMIC LANGUAGE ALGOL 60
uses the spelling "ALGOL" throughout the body of the text, *but* also
says
	"This conference also led to the publication by Regnecentralen,
	Copengagen, of an {\em Algol Bulletin}, edited by {\sc Peter Naur}."
where Peter Naur's name is in capitals and Algol is not.  And the copy I
just checked is in a book called
	A Course in Algol Programming.
The
	Report on the Algorithmic Language ALGOL 68
uses capitals, but 
	Algol 68
	a first and second course
by McGettrick has mixed case on the cover (caps inside).

The evidence is that "Algol" was regarded as acceptable by many of
the people in the Algol community.  The "Algol Bulletin" was _the_
discussion organ of that community.

	BASIC is an acronym.

Beginners' All-purpose Symbolic Instruction Code.  Indeed.

	FORTRAN is an acronym.

Not in the relevant sense.  "Formula Translation".  If it were FT,
the usage argument would have some force.  Remember "Comintern".

	LISP is an acronym.

"LISt Processing".  Note the point of the capital letters convention:
the letters of an acronym (strong sense only) are capitalised because
we capitalise the first letter of words.  Lisp is only an acronym in
the weak sense, and the all-caps convention has never been required
for words like that.

It is important to understand that programming language names were
capitalised because they were programming language names.  The programming
languages Euler, Pascal, and Euclid were and are often written EULER,
PASCAL, and EUCLID, although they are not acronyms in any sense at all.

	PL/1 is an initialism.

No, PL/1 is a solecism.  It's PL/I with an eye, not a one.
Programming Language/One.

	APL is an initialism.  BAL is an initialism.  SQL is an initialism.
And as such irrelevant.
	
	Ada is not an acronym.  Java is not an acronym.  Pascal is not an acronym.

However, you will *very* often find ADA and PASCAL written.
For example, in Harel's book "Algorithmics: the Spirit of Computing",
we find in the index on page 418 ADA, MODULA-2, OCCAM, PASCAL, SIMULA-67
among others.  I suppose you could argue for "MODUlar LAnguage" and
"SIMUlation LAnguage", but Occam is a name just like Pascal.

	Smalltalk is not an acronym.
	
Nor are Ada, Occam, and Pascal.  The Smalltalk programming language is
spelled "Smalltalk" because that's how the designers said it was spelled
and that's how they still say it's spelled.

	Programming Jeopardy
	Languages for $100: FORTRAN.
	Answer: "What For?"
	
If you want
 - extremely efficient code
 - modularity
 - typesafe overloading
 - user defined data types
 - arrays with dynamic sizes
 - nested procedures
 - a reasonable range of control structures
 - a standard interface to UNIX
 ...	
then modern Fortran isn't bad at all.

I think that little question and answer perfectly demonstrates my point
about the spelling indicating whether one has kept up to date.




More information about the Squeak-dev mailing list