Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site fortune.UUCP
Path: utzoo!linus!security!genrad!grkermit!masscomp!clyde!ihnp4!fortune!rpw3
From: rpw3@fortune.UUCP
Newsgroups: net.lang
Subject: Re: Strong Typing and Ignorance - (nf)
Message-ID: <1968@fortune.UUCP>
Date: Wed, 14-Dec-83 07:04:19 EST
Article-I.D.: fortune.1968
Posted: Wed Dec 14 07:04:19 1983
Date-Received: Fri, 16-Dec-83 01:48:01 EST
Sender: notes@fortune.UUCP
Organization: Fortune Systems, Redwood City, CA
Lines: 27

#R:utcsrgv:-289200:fortune:15100004:000:1017
fortune!rpw3    Dec 14 03:37:00 1983

When people called BLISS "un-typed", what they really meant was that
the contents of a field of storage was uninterpreted by the compiler.

In fact, while (simple) variables were untyped, the OPERATORS were
typed. There was integer add ("+"), and floating point add ("FADR"), etc.
No coercion. If you wanted to write it (and since BLISS was for
writing operating systems and libraries, you sometimes did), you could:

	x = (.y FADR .z) + 1^27 ;

which takes the contents of "y" and the contents of "z", adds them with
floating-point arithmetic, and then integer adds 2**27 (1<<27, in C) to
the result (bumping the exponent, on a PDP-10), and store it in "x".

"Un-typed" generally means "no type checking", although in some contexts,
such as APL and SNOBOL, it means self-described typing (descripter-based
variables). Is that sufficiently muddy?

Rob Warnock

UUCP:	{sri-unix,amd70,hpda,harpo,ihnp4,allegra}!fortune!rpw3
DDD:	(415)595-8444
USPS:	Fortune Systems Corp, 101 Twin Dolphins Drive, Redwood City, CA 94065