Subject: Re: "k"bytes or "K"bytes?
To: der Mouse <mouse@Rodents.Montreal.QC.CA>
From: Sean Doran <smd@ebone.net>
List: tech-kern
Date: 10/26/1999 21:03:18
der Mouse  <mouse@Rodents.Montreal.QC.CA> writes:

> It's not as if they don't *know* they're being misleading; I've even
> seen footnotes in some ads recently saying things like "based on 1MB =
> 1 million bytes", which I take as clear evidence that they *know*
> they're being misleading and that people mind.

I think that the people who mind is a proper subset of the
people who know the 1Mbit = 2^10 bits convention.  

I resent the tyranny of addressing lines over descriptons
of data volume.

(It's particularly annoying to when someone starts talkling
 about how long it takes a network with an end-to-end goodput
 of 100 Mbits/second to transmit 100 MBytes of data.
 It's not eight seconds, and that's not just because a
 byte is not unambiguously 8 bits.)

I have some sympathy for the OSI people who tried to sort
this out with the term "octets", where a Mega-octet would
8 * 10^6 bits, whether talking about transmission speed or
data volume.

> If only we could get a sympathetic memory company to make SIMMs holding
> exactly (say) 32,000,000 bytes, and whenever they get an order from a
> disk maker, send out those instead of the regular sort... >:->

Personally I would rather get a sympathetic memory company
that describes a SIMM as containing 2^a * 2^d bits of
addressable memory, where 'a' and 'd' are respectively the
number of addressing lines and data lines the SIMM pays
attention to.

        Sean.