Re: Bad Metric ... 1k = 1024 or 1000?

[wbdesnoy@acs2.acs.ucalgary.ca (Byron Desnoyers Winmill)]

Supertimer (supertimer@aol.com) wrote:
: Masterkit <masterkit@mediaone.net> wrote:
: >For anyone who really thinks a k is 1000 bytes, you're wasting your time on
: >them.  They're decimal-only types.
: far
: It is easy to dismiss the matter as ignorance, but the fact is
: a "kilo" designation is defined as a "1000x" metric designation.
: The whole purpose of metric is to standardize measures.

The problem is that metric does not use bit or bytes as a unit of
measurement either.  (It is rather disturbing that bits and bytes
are both represented by "b".)  In otherwords, it is computer users
who accepted kilo and mega as a matter of convenience -- rather than
being bound to it by some standard.

: Then why did the computer folks start using the kilo
: designation in the first place?  This is not a "well they
: don't know a thing about binary or hex" matter.  It is a
: matter of definition.  Kilo means one thousand of
: something.

In my mind, it is a matter of convenience. 2**10 is awfully close
to 10**3, so that may as well be the prefix.  Just imagine getting
one from the other by ror (or is it rol?) ten times.  Elegant.
Certainly more creative than the normal sort of division.

But you are right: one way or the other it must be cleared up.
Now that we are getting into gigabyte media, using 1000 instead of
1024 is a good way to pull the wool over the consumer's eye: in
kb, it made about a 2% difference (so what!), then about 4% for
Mb, now about 7% for Gb.

Perhaps it should be 1000 to agree with metric, or perhaps it should
be 1024 to keep memory addressing clean and memory sizes memorable
(would you prefer 64Mb or 67.11Mb?)

Well, it is not too much of a bother for myself nowdays.  The difference
for the kb and Mb range is sufficently small that it does not matter.
And kb and Mb are what matters for the IIgs.

--
Stopped (SIGTTOU)         byron.
$