final sigma and tonos

[Stephane Bortzmeyer]

On Thu, Jan 31, 2008 at 04:42:35PM -0800,
 Erik van der Poel <erikv at google.com> wrote 
 a message of 37 lines which said:

> > But someone could come forward and say that for French
> > we leave off the accents on uppercase vowels. And
> > because of that:
> >
> > école.fr should (but won't) match ECOLE.fr as IDNs.
...
> You are a respected member of the Unicode community. Nevertheless, I
> would also be quite interested to hear how the Greeks and French on
> this mailing list feel.

I certainly cannot speak for all the French people but here are my two
eurocents:

* the uppercase of é has always been É (with the accent), not E

* during the typewriting era, most keyboards did not offer an easy way
to type accented uppercase characters. As a result, some people
started to drop the accents on these and it started a legend that "In
French, you do not put accents on uppercase characters" (this is
false, but, as every linguist knows, the typos of today may be the
features of tomorrow).

* when the computers replaced the typewriters, the trend was
reversed. 

* you can see the historical evolution in cheap books (expensive books
were always right): if it has been printed between the 1920s and the
1990s, there is a good chance it has no accented uppercase character.

As of today, the issue is still disputed. Typically, for matching,
many French speakers assume that matching will be accent-independent
and the proposed - not implemented in ".fr" - RFC 4290 table in
<http://www.bortzmeyer.org/4290.html> (the article is in French but
the comments in the table are in English) therefore create bundles
that way.

Warning: this is for the French language, as spoken in France (I've
been told it was different in Québec), not for all European languages
with accents. Speakers of these languages have different feelings
about accents on uppercase characters.