3-in-1

From: "David Wilson" <davidwwilson@comcast.net>

Back in 1977, when I first set my hands on a computer, a gigabyte of storage would have been a phenominal luxury. Today, it is as common as dirt to hear "gig of disk." Most computer users see a "gig" as a unit of disk space or memory, much as a cup is a unit of liquid. They neither know nor care that "giga" means 10^9, that a "gig" or gigabyte is actually 2^30 = 1,073,741,824 bytes, or

This is a common misconception. It has only ever been true for RAM where the technology itself has been physically growing by powers of two. Ah, Microsoft has propagated it too, and it's also been true in the uninformed media, alas, hence the commonness of the misconception. In telecommunications, they've been working with 1000 octets a second for their kb/s, and the units multipliers all scales by powers of 10. E.g. my Nokia BB2M is rated as 2.048Mb/s, meaning 2048000b/s, to match G.702, and the entire plesiochronous digital hierarchy (US Sonet) is based on multiples of that measurement. SDH also followed the same well-entrenched tradition as PDH. The use of multipliers as powers of ten in communications to measure large quantities of data predates other computer uses of the term. Hard disk technology is where the real confusion arises. Hard disks simply aren't limited by the same power of two increases as RAM technology, and hard disks have had arbitrary sizes. Most, but not all, hard disk manufacturers have always described the size of their disks using decimal powers, rather than binary ones. The popular media has incorrectly called this deceitful, because their OS choses, equally arbitrarily, to divide by powers of 2 and report a smaller size. Considering the raw data capacity of a hard disk has _never_ been reflected by the formatted capacity of same, due to boot sectors, partition tables, FATs, superblocks, reserved space, and simple cluster/sector wastage, the popular press' arguments are mostly without merit. Why do they not complain at the OS for chosing the 2^20 scaling factor rather than the HD manufactors for using the 10^6 one? To guarantee confusion, there _is_ a real reason, alas. It's basically that hard disks have had power-of-2 based sector sizes - typically 512 bytes, and so there has been _some_ binary tradition. Recently, 1996 IIRC, some utterly horrible prefixes were invented as an attempt to reduce the confusion. Pedantically, _only_ Gibi is 2^30 nowadays. This keeps the telecomms guys and the hard disk guys happy, at least, as the sanity of ISO 31 is restored at least on paper.

Allan Wechsler wrote:

...

Kilo, Mega, Giga, Tera, Peta, Exa, Heta, Oto, Nea

...

Starting from Tera, they are derived by doing some sort of surgery on the Greek word for the log base-1000. For Tera, Peta, and Exa, I believe this is just a coincidence; the idea behind the extension to Heta, Oto, and Nea is to pretend that the coincidence was intentional.

Indeed. The etymology I have for tera is Gk. teras, monster (giga was Gk. gigas, giant of course). The term was coined in 1947 by the UIC. When a prefix for a larger multiplier was required it has been suggested, but not verified, that in 1975 the CIPM created the pretension tera <- tetra, and extended it to peta <- penta and exa <- hexa. If the full OED can't be sure of this derivation, I fear only the people who were in that room in 1975 can. Some have claimed that it was the CGPM who coined the terms, but the CGPM sat in December, and the CIPM sat in March. The reducing multipliers were given their scandinavian numerically-based prefixes, femto- and atto-, much earlier; in 1961 by the IUPAP. (pico- being nearly a hundred years old now, and not number-based) And back to the subject line: <<< In 1993, hacker Morgan Burke proposed, to general approval on Usenet, the following additional prefixes: groucho (10^-30), harpo (10^-27), harpi (10^27), grouchi (10^30). This would leave the prefixes zeppo-, gummo-, and chico- available for future expansion. Sadly, there is little immediate prospect that Mr. Burke's eminently sensible proposal will be ratified.

...

...

http://burks.brighton.ac.uk/burks/foldoc/44/92.htm

And the somewhat joking, but in some ways reasonably soundly justified: <<< xona, weka, vunda, uda, treda, sorta...

...

...

http://jimvb.home.mindspring.com/unitsystem.htm (Basically, scan back through the alphabet for the first character, and make the rest of the prefix resemble a greco- latin word corresponding to the power of 1000, such as x+nono->xono, w+deka->weka, etc.)

It's in the hands of the standards bodies. If gibi and yotta are possible, then pretty much anything is! Phil ===== When inserting a CD, hold down shift to stop the AutoRun feature In the Device Manager, disable the SbcpHid device. http://www.cs.princeton.edu/~jhalderm/cd3/ __________________________________ Do you Yahoo!? Vote for the stars of Yahoo!'s next ad campaign! http://advision.webevents.yahoo.com/yahoo/votelifeengine/