On Tue, 13 Mar 2018 16:48:51 -0700, Asmus Freytag (c) via Unicode wrote:

On 3/13/2018 12:55 PM, Philippe Verdy wrote:

It is then a version of the matching standards from Canadian and French 
standard bodies. This does not make a big difference, except that those 
national standards (last editions in 2003) are not kept in sync with evolutions 
of the ISO/IEC standard. So it can be said that this was a version for the 2003 
version of the ISO/IEC standard, supported and sponsored by some of their 
national members.

There is a way to transpose international standards to national standards, but 
they then pick up a new designation, e.g. ANSI for US or DIN for German or EN 
for European Norm.


2018-03-13 19:38 GMT+01:00 Asmus Freytag via Unicode :

On 3/13/2018 11:20 AM, Marcel Schneider via Unicode wrote:

On Mon, 12 Mar 2018 14:55:28 +0000, Michel Suignard wrote:

Time to correct some facts.
The French version of ISO/IEC 10646 (2003 version) were done in a separate 
effort by Canada and France NBs and not within SC2 proper. 

Then it can be referred to as “French version of ISO/IEC 10646” but I’ve got 
Andrew’s point, too.

Correction: if a project is not carried out by SC2 (the proper ISO/IEC 
subcommittee) then it is not a "version" of the ISO/IEC standard.


Thanks for correction. And I confess and apologize that on Patrick’s French 
Unicode 5.0 Code Charts page (
), there is no instance of "version", although the item is referred to as "ISO 
10646:2003 (F)", from which it can ordinarily be inferred that "ISO" did back 
the project and that it is considered as the French version of the standard.
I wasn’t aware that this kind of parsing the facts is somewhat informal and 
shouldn’t be handled on mailing lists without a caveat. 
That said, the French transposition of ISO/IEC 10646 was not carried out as 
just sort of a joint venture of Canada and France (which btw has stepped out, 
leaving Québec alone supporting the cost of future editions! Really ugly), 
given that it got feedback from numerous countries, part of which was written 
in French, and went through a heavy ballot process. Thus, getting it changed is 
not easy since it was approved by the time, and any change requests should be 
documented and are primarily damageable as threatening stability. Name changes 
affecting rare characters prove to be feasible, while on the other hand, 
syncing the French name of U+202F with common practice and TUS is obviously 
more complicated, which in turn compromises usability in UIs, where we’re 
therefore likely to use descriptors i.e. altered names for roughly half of the 
characters bearing a specific name. Somehow the same rationale as for UTN #24 
but somewhat less apposite given that the French transposition is not 
constrained by stability policies.
Best regards,

Reply via email to