> Because the binary multipliers are used in very limited circumstances,
> and where the actual number doesn't really matter, and you don't have
> any choice.  I mean, suppose a computer has 512MiByte of memory.  Does
> the consumer care that it's more than 512Mbyte?

You may have got a point here. But I have read complaints about disk size
comparisons (GB vs GiB) from consumers, so they do seem to care. (Of course
they didn't talk about GiB, but they complained that one company announced
disk sizes in M=10^6 and the other in M=2^20. I think pretty much all disk
manufacturers now add some small print where they state what their specific
M or G or T is supposed to mean. Now /that's/ a waste... wouldn't it be
easier to simply use unambiguous units in the first place?

> Are there applications that
> would work in 512Mibyte but not 512byte?

Probably, but I take it that's not what you meant :)

> Are binary multipliers used ANYWHERE other than memory size these days?
> Are they ever likely to be used for anything else?

Yes, in a way. Related to memory size is amount of data. They are commonly
used to denominate amounts of data ("this program has 2.5 kB"), and speed
of data transmission ("my average download speed is 32.4 kB/s"). Now you
can say that's all memory size, because all that data ends up or passes
through memory. But you also can say that I'm not talking about memory
here, I'm talking about data.

I definitely think that it can make a difference whether I'm talking about
2.5 kB or 2.5 kiB. The former may fit where the latter doesn't. Now here
you could say that we are back to memory size, and that memory size uses
the binary multipliers. But how does that match with the fact that most
disk manufacturers use decimal multipliers?

Then there are the transmission speeds. Traditionally, the communications
guys used to look at individual bits and how many per second, and relate
that more to the 10-based frequency prefixes, so if you read 3 Mb/s, that's
likely to be 3 * 10^6 bit per second. But then the more program size
oriented programmers started to transmit data, and they said that they are
transmitting 3 MB/s, and this is most likely 3 * 2^20 byte per second. But
neither is guaranteed, as the former could have been reached at by
multiplying the 2-based 375 kB/s by 8, and the second could have been
reached at by dividing the 10-based 24 Mb/s by 8. So you never know... 4.8%
(M vs Mi) or 7.3% (G vs Gi) can make a difference. This can eat up all the
safety you built in, or more.

I agree that in many cases those it doesn't make a difference, but I think
using clear terminology in technology is better -- and not more effort.
That's the "MA" vs "mA" thing... I prefer "mA" because it doesn't cost me
more to use a clearly defined standard than it would cost me to use an
ambiguous unit ("MA") that is only useful within a small domain.

This is all about communication. In the lone rider setup, it's likely not
to make much of a difference. But in a collaboration situation, it's really
good to know what everybody's M means, and to be able to rely on the
information one receives from the others. To the point, not within give or
take 10%.

(BTW, consumers don't really bother me much in this context. This is more
about the people who know the difference.)

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads