CONVERT - nonstandard prefix "e" instead of SI standard prefix "da"

["David A. Wheeler" <dwheeler@dwheeler.com>]

Okay, this CONVERT function is just downright bizarre, I'm glad I'm going over this thing carefully. It turns out that Excel/Ecma use the non-standard prefix "e" instead of the SI standard prefix "da". Many people are going to HATE that; I suggest we accept BOTH "da" AND "e".

Background: One of the SI prefixes is "da" (pronounced deka), value of 10^1.  It's not as common as many other prefixes, but it's been a part of SI from the beginning.  This is NOT the same as "d" (pronounced deci), which has the value 10^-1.

In Excel's CONVERT, the standard "da" is not used... instead, we have only "e".  I  just did a check, and there's no name conflict when "da" is accepted as a prefix as well as "e".  Yes, this means you have to check two characters instead of just one to examine a prefix, but binary prefixes require that anyway (Mi, etc.).

I suggest that we accept BOTH "da" (correct) or "e" (historic), and then alert ISO that the Ecma spec has yet another standards conflict - in this case, a conflict with the SI prefixes.  We may as well also notify them that it also fails to accept "s" for second, which is the standard abbreviation for that base unit in SI (I believe we should accept both "s" and "sec" for the same reasons).  If you think that's a bad idea, let me know; the same arguments would apply to both.

--- David A. Wheeler