Yes, I did take note of that. However, the logic would be: If 'b' is for byte, what symbol would we be using for a bit?
I have no knowledge of the historical background of the uppercase B. However, quoting from http://en.wikipedia.org/wiki/Byte : "Byte" is most often abbreviated as "B", hence "MB for "megabyte". This is incompatible with the SI system of units, because "B" is already used to represent the Bel. Sometimes "b" is used for byte (and "bit" for bit), but this can cause confusion because of the much more common use of "b" to mean bit." It's all a mess. Neither 'b' or 'B' may actually be used to inambigously represent the byte, and neither 'k' or 'K' should be used for "kilo". ('b' is ambigous: It means both byte and bit (and sometimes even baud). 'B' is conflicting with the Système Internationale. 'k' actually means 10^3. And, 'K' has some taste of ambigouty to it, too.) I believe it becomes a question of opinion, then. When it comes to the prefixes, I personally think that "kilo" has always been 'k', and that if it should be traded for anything, it should be "Ki". And perhaps, Debian is one of the better starting points for this trend? About the units, I feel for 'B', as 'b' is often used for bit and baud. This is knowingly breaking the SI, I know. But -- Industries have had a tendency to derive defacto units from the SI, which fits their needs. And by the way, I apologize, if my last post was rude. Reading through it myself, I realized that it might actually be just that. Regards, Anders Breindahl.
