NetNews Usenet Archive 1992 #30/NN_1992_30.iso/spool/comp/arch/11700

Internet Message Format | 1992-12-16 | 4.0 KB

Path: sparky!uunet!stanford.edu!snorkelwacker.mit.edu!ai-lab!life.ai.mit.edu!burley
From: [e-mail] (Craig Burley)
Newsgroups: comp.arch
Subject: Re: Date request
Date: 16 Dec 92 13:53:11
Organization: Free Software Foundation 545 Tech Square Cambridge, MA 02139
Lines: 65
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]>
    <[e-mail]> <[e-mail]>
NNTP-Posting-Host: apple-gunkies.gnu.ai.mit.edu
In-reply-to: [e-mail]'s message of 16 Dec 92 16:11:33 GMT

In article <[e-mail]> [e-mail] writes:

   In article <[e-mail]>, [e-mail] (Bohdan Tashchuk) writes:
   > A 10/21/92 semi-official net posting on NextStep claims it runs on SXs:

   The date above refers to 21 November 1992. That may seem obvious to some of
   you.

Well, to me, it's 21 October 1992.  Do you Brits start numbering months
at 0?  :-)

   It would be more clear if everyone wrote the month name as a name (until
   2001 anyway). ie: 21/NOV/92.

I liked the DEC convention of 21-NOV-92 or 21-Nov-92, but I think the
planet is best served with the simplicity of YYMMDD or YY/MM/DD for clarity.
Until 2001, it'll be very obvious that one is using that notation since
MM and DD never venture into the 40s+ realm.  And "Nov" doesn't mean
"November" in every language, or even in some quite-popular languages
used in "first-world" countries.

   The above aside, I'm always puzzled as to how the American date convention
   started: Mont/Day/Year is neither LSB or MSB, where as Day/Month/Year and
   Year/Month/Day have obvious reasoning behind them.

Probably because we've historically "said" dates as "October 3, 1992", so
shortening that becomes 10/3/92.

Why you expect "obvious reasoning" from a U.S. system of measurement is
beyond me, however...we're still using gallons, feet, and pounds to
measure many things here, and, unbelievably, many people here think
those are _easier_ to deal with than the Metric System.  I (and many others)
are more _used_ to our current system of measurement, but devoting
about 10 minutes to issues Metric is, I find, enough to understand it,
compared to lots of memorization for our current system.  And it lends
itself to much more enjoyable abuse because of the flexibility of
the prefix/postfix design -- what's the equivalent of a "microprocessor"
in the English system, an "inch-processor"??  What's the equivalent
of "nanotechnology"?  Metric is superior to the US system (which I hesitate
to call the English system, though I think that's technically correct,
so I don't offend anyone) in about the same way that English is superior
to that Eskimo language that has five words for "snow" -- the latter has
more specialization, but less flexibility in dealing with new concepts
(specifically, with technological progress), something that is and
will continue to be a fact of life for at least another century.

At the very least, people on the net should try and use more international
ways to represent things when they can, and with dates, that's especially easy.

So I suggest we use YYMMDD or YY/MM/DD whenever possible, since that seems
to be the most universally understood, even if it isn't the most accepted
form in any given culture.  (Not everyone numbers years "A.D.".)

But then I always thought it was clearly best to label bit numbers in an
architecture manual (and machine design, where appropriate) starting with
0 for the least-significant bit and going up from there, so the integer
value of a given bit is 2^n, where n is the bit number.  And it always
surprised me to find how many architectures were _not_ done this way!

(The above paragraph is not necessarily pertinent to the discussion at
hand, but it is probably so for comp.arch.  :-)
--

James Craig Burley, Software Craftsperson    [e-mail]
Member of the League for Programming Freedom (LPF) [e-mail]