Newsgroups: sci.lang,alt.usage.english
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!scramble.lm.com!news.math.psu.edu!news.cac.psu.edu!howland.reston.ans.net!newsfeed.internetmci.com!EU.net!usenet2.news.uk.psi.net!uknet!newsfeed.ed.ac.uk!dcs.ed.ac.uk!faither.dcs.ed.ac.uk!rwt
From: rwt@dcs.ed.ac.uk (Rainer Thonnes)
Subject: Re: kByte, KByte, Kbyte, kbyte???
X-Nntp-Posting-Host: faither.dcs.ed.ac.uk
Message-ID: <Dr6wty.8sn.0.staffin.dcs.ed.ac.uk@dcs.ed.ac.uk>
Sender: rwt@faither.dcs.ed.ac.uk (Rainer Thonnes)
Organization: Edinburgh University Computer Science Department
X-Newsreader: xrn 8.00
References: <317B8356.40E9@ov.fr.bosch.de> <4m28li$69o$1@sydney.DIALix.oz.au> <ADAD698D9668D90B3@hearsay.demon.co.uk> <Dqu9Lx.BDw.0.staffin.dcs.ed.ac.uk@dcs.ed.ac.uk> <chris-c-0905960149570001@gw3-019.pool.dircon.co.uk>
Date: Fri, 10 May 1996 12:46:43 GMT
Lines: 29

In article <chris-c-0905960149570001@gw3-019.pool.dircon.co.uk>,
chris-c@dircon.co.uk (Chris) writes:
> rwt@dcs.ed.ac.uk (Rainer Thonnes) wrote:
> 
> > Interestingly enough, 'k' seems to be the only exception to the rule you
> > alluded to, which uses capital letters for multipliers greater than one
> > and lower-case ones for those less than 1.
> 
> I hesitate to join this furious argument, but I was taught at school
> exactly the same rule - multipliers are capitalised, dividers are lower
> case, e.g. 'm' is 'milli', 'M' is Mega.

I was convinced, but mistakenly so, that "hecto" became 'H', and that "hPa"
was therefore incorrect.  Alas, it really is 'h', so now there are two
exceptions.

Perhaps the real rule is that whenever the *same* letter is used for both
positive and negative powers of ten, the positive one uses the capitalised
one.  Hence m/M (milli/mega) and d/D (deci/deca).

Sadly, that rule isn't very useful, since it applies only to a universe of
two cases.  It's strengthened a little, by the fact that most other
multipliers seem to fit in.  Unfortunately, as soon as you start to widen
the scope of a rule, you have to admit exceptions.

The most important rule is that most rules have exceptions.
Another oddball is "da" as an alternative to 'D'.

afpnum(cd)/(Dh)kMGT  -  can anyone extend this?
