In a message dated 08-25-99, Jonathan de Boyne Pollard said to Holger
Granholm:

 JP>> "1.5Ki" is "1.5 kilobinary".  But the obvious question in the
 JP>> reader's mind, if that *had* been what I had written, would be "1.5
 JP>> kilobinary *what* ?".  The answer is "1.5 kilobinary *bytes*", which
 JP>> is, of course, abbreviated to "1.5KiB".

What however was/is confusing to me is your expression "KiB". According
to the information I have, there is no abbreviation like that.

JP>For anyone with any experience of computers (which certainly covers
JP>both you, me, and probably anyone reading this echo!), multiples of
JP>1024 are hardly something "new" that has to "sink in".  

Agreed.

JP>"kilobinary-", or "kibi-", simply means "1024", in the same way that
JP>"kilo-" means "1000".  It's very logical and straightforward as far
JP>as I can see. Neither of them has anything to do with bytes.

Yes, and as 1024 is a binary number I find it confusing to add the "B"
to Ki which in itself already expresses that we are talking about binary
numbers, ie. bytes not bits.

JP>They are a very good one.  They are so sensible and straightforward
JP>that I started using them almost immediately after I read about
JP>them.  

Me too, until I saw your interpretation. Maybe what I read about the new
memory standard wasn't complete.

Have a nice day,

Holger

___
 * MR/2 2.26 * Windws is ine for bckgroun comunicaions


--- PCBoard (R) v15.22 (OS/2) 2
102