Xref: utzoo comp.lang.c:12534 comp.arch:6322
Checksum: 42911
Lines: 22
Path: utzoo!sq!msb
From: msb@sq.uucp (Mark Brader)
Date: Tue, 13-Sep-88 18:26:09 EDT
Message-ID: <1988Sep13.182609.230@sq.uucp>
Newsgroups: comp.lang.c,comp.arch
Subject: Re: non-binary hardware
Summary: goto comp.arch;
References: <1285@mcgill-vision.UUCP> <3473@phri.UUCP> <1988Sep12.032804.25762@utzoo.uucp>
Reply-To: msb@sq.com (Mark Brader)
Followup-To: comp.arch
Organization: SoftQuad Inc., Toronto

[1] The word is "ternary", not "trinary".   If you want to use the "tri-"
    prefix, you have to say "triadic".  Of course, as with all things in
    language, this is subject to change if "trinary" becomes sufficiently
    popular; the words for bases 8 and 16 both changed when they became popular
    with computers.  (See Knuth, volume 2, sec 4.1; in 1st edition, p. 168.)
    But "ternary" is the accepted form.

[2] Please move this discussion out of comp.lang.c.  Discussion of non-
    binary machines should probably go in comp.arch (I'm cross-posting to
    there and directing followups to it).  If discussion of terminology
    occurs, it should probably go to sci.lang.  Sci.math is another possible
    group for some aspects.

[3] While I'm posting, I may as well point out that the ENIAC calculator
    (or computer, depending on your definition; it was plugboard-programmed)
    of 1945 used decimal, non-BCD arithmetic, but the underlying implement-
    ation was binary; each of the 10 digits in each of its 20 registers
    contained 10 flip-flops each containing 2 vacuum tubes!

Mark Brader			"You can't [compare] computer memory and recall
SoftQuad Inc., Toronto		 with human memory and recall.  It's comparing
utzoo!sq!msb, msb@sq.com	 apples and bicycles."		-- Ed Knowles