Mr Schwarz's comments are a mixed bag. His conjectured typo was a typo, and it
would be rude for me to comment further on the paucity of his vocabularity, of
which I should in any case need a larger sample to make a judgment.
DSN is a systematically ambiguous acronym. In the context in which I used it
it denotes Day Serial Number; in an MVS data-management context it denotes
DataSet Name; in at least one bird-watching context it denotes Daily Sightings
News; etc., etc.
Are there dates more than 5 million years in the future or past that can be
known with great precision. The answer is yes, at least for such astronomical
event as eclipses, which can be known outside this interval to the hour and
minute. For many other events such precision would be delusive, but even
approximations must often be used in arithmetic expressions in which values
that are known with great precision also figure, and for this purpose they must
be expressed in the same way.
An example will; help here. Suppose that a news service reports a bank raid in
Nottingham that netted the miscreants £5000. It may then be reported an
American newspaper as a bank robbery that netted them $7774, the spurious
precision of this second result having resulted from the use of the [current]
exchange rate £1 = US$1.5548.
We mostly lack conventions for dealing with such issues, chiefly because they
are difficult to explain to the untutored. [There are, of course, contexts in
which we do have such conventions. Consider two of Cantor's transfinite
numbers, T and t, T > t. Then T + t = T. The sum of two transfinite numbers
is the larger, and in general sum(t(i), i = 1, 2, . . . , N) = max(t(i)).
Regrettably, These and similar devices need too much explication to be at all
readily portable.]
The question whether zero-origin or one-origin values should be used in some
context is an old one. The traditional Gregorian year progression is still
. . . ,2 BC, 1 BC, AD 1, AD 2, . . .
It goes from 1 BC to AD 1 without passing through a zeroth year or collecting
$200. The astronomical Julian calendar, on the other hand, has a zeroth year.
Or again COBOL uses one-origin subscripting for tables; and C uses zero-origin
subscripting for arrays. Asked about this difference, Admiral Hopper, who was
as mathematician, said that she used one-origin tables when talking to
accountants and zero-origin arrays when talking to her quondam Harvard
colleagues. Context is all!
All this is by the way. Mr Schwarz wants me to write my posts in a species of
subliterate English he judges appropriate. I demur, as I have in the past;
and, as I have noted in the past too, he has an immediate remedy at hand. He
can put my email address on his kill list, which will make me an unperson to
him and spare him the sight of my posts.
John Gilmore Ashland, MA 01721-1817 USA
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
