> Except that these problems are at a higher level, when > dealing with  words.  Strings really just deal with > characters and have no idea  about languages.

OK, so why not make that the next level of abstraction, and leave that to the next generation working on the problem?  In Squeak, strings also deal primarily with characters and have no (meaningful) understanding about langauges (other than parsing aids for the compiler).

> If it helps, I can probably provide class-documentation of > the  NeXT/Apple NSString class-cluster, which does exactly > that,  successfully.

A model would be a great thing to consider.  Here's one mile-long URL to a breakout of NSSString:

http://developer.apple.com/techpubs/macosx/System/Library/Frameworks/Foundatio
n.framework/Versions/C/Resources/English.lproj/Documentation/Reference/ObjC_cl
assic/Classes/NSString.html

Is there a more concise breakout anywhere, or do you suggest we start there?  I for one, think this can be a fairly straightforward exercise if we do it right and try only to build the simplest thing that could possibly work.  It would be better than the status quo, address most real issues presently at hand, and provide greater flexibility for the future.

Anyone else game to play?