From: utzoo!watmath!idallen
Newsgroups: net.works
Title: Re: Bandwidth, encodings, abbreviations and command language.
Article-I.D.: watmath.3130
Posted: Fri Jul 30 00:35:08 1982
Received: Fri Jul 30 01:48:55 1982
References: populi.261

It was said (populi.261 from G:wing) that VAX/VMS is a very human
command language.  The text of the article went on to state that all
VMS commands were words, and abbreviations just abbreviated each word.

This may be true, but if this article was meant to be a followup to my
article on bandwidth and encoding, then the point of my news item was
missed.  (I do try to make myself clear, but it's not easy!)

The followup article might reasonably argue that VMS has a human-oriented
*abbreviation scheme*; but, the article does not give any evidence that
the unabbreviated *language itself* is human-oriented.

My main point was that we mistake bandwidth-improving encoding schemes
(abbreviation is one such scheme) as being *language design*.  Everyone
is saying how wonderful their abbreviation scheme is; nobody is saying
how wonderful their *command language design* is.  I don't much care how
easy it is to abbreviate the EXPUNGE_DATASET_WITH_QUERY command, if I
can't even remember that it is the command I have to use.

I strongly believe that the *first* requirement for command language
design is to make the language *memorable*.  Once we remember *what*
we want to say, we can, at our leisure, learn encoding (abbreviation)
schemes to speed things up at the keyboard (or mouse, or ...).

	-IAN!    U of Waterloo    (decvax!watmath!idallen)