At 12:19 PM 3/29/00 -0800, you wrote:

>From: David Schnepper <dschnepper@...>
>
>
>(By the way - this points out the need to have ACCENT-Insensitive collation
>orders, as well as
>case-insensitive orders).
>

Sigh. I have to say I expected this.

Dave (and others): How many other close relatives does CASE/NOCASE
have? How many classes of collation classes should be expect?
How many flavors of case insensitivity are the French likely to
have?

Another question: Do you know of any cases where a "case neutral"
representation from equivalent upper and lower cases?

I know you are familiar with the famous Knuth article on collation.
I don't think we need to collate "1812 Overture" as the letters
forming the German words to 1812, but leaving room clearly useful
non-English equivalents is important.

Does the collation driver mechanism currently support case changing?
If so, can a single pair of entrypoints (upcase, downcase) suffice.
If the French want separate ACCENT/NOACCENT collating sequence, can
we reasonable expect them to support two different drivers?

How much trouble are we getting ourselves into?

Jim Starkey