I noticed a while ago that Azurius reports data sizes/speeds in Mebibytes (MiB) instead of Megabytes (MB). What other clients use the new terminology, or at least use the words "mega" and "kilo" [correctly] according to the metric system? I was reminded of this because I was trying to make some size-filters in gtk-gnutella, and had a difficult time because of the program's casual use of the abbreviation MB to mean 2^20. Anybody else have an opinion about this admitedly pedantic issue?

I don't know much about it but I thought only hardware manufacturers used MB for marketing purposes. And I thought all softw based measurements were in the 'TRUE' measurements. ie: one megabyte 1 MB = 106 B = 1 000 000 B
one mebibyte 1 MiB = 220 B = 1 048 576 B

So in that light I thought you should be using mebibytes or its equivalent. I'm surprised Gnutella uses the other!

But then again, I suspect on the mac there's a lot of rounding off that goes on.

I really don't think it's a platform issue. I'm on SuSe Linux. I think it is merely a very widespread bad habit among programmers. I have seen lots of software that uses the abbreviation MB to signify 2^20 and KB to signify 1024.

Well I was never even aware of the term MiB or the others even in my training (unless I missed something lol). We were just told that there were 2 techniques used for calculating them; one correctly & the other not. So I guess in that context .. one is wrong. lol I thought the so-called KB did signify 1024 on OS systems. Which is why after formatting a HDD it gives less than the manufacuter's designated size. I think I'm lost lol.