AW: AW: sharp s (Eszett)
fsasaki at w3.org
Fri Mar 14 02:58:16 CET 2008
Martin Duerst wrote:
> At 08:36 08/03/12, Felix Sasaki wrote:
>> John C Klensin wrote:
>>> --On Tuesday, 11 March, 2008 12:12 -0700 Mark Davis
>>> <mark.davis at icu-project.org> wrote:
>>> While it is clear that it would be easy to say "we have to stick
>>> with Unicode casefolding rules" or "compatibility with IDNA2003
>>> is more important than anything else", I'm not comfortable
>>> telling the users that we've decided that they don't get to use
>>> this character because it is inconvenient.
>> The problem of this process is to find out who represents "the user". Is Georg representing the German language user community, or is the input of this community to Unicode a better representation, or ...?
> I very much think this is a question of what users need. If I had
> enough German speakers nearby, I'd design some experiments.
> But I think it's a bad idea to play different inputs against each other.
agree, that was not my intent. Sorry if it appeared that way, and I'm
very happy to see Georg's input.
> The Unicode table was created mainly for search purposes, where false
> positives are less of a problem than false negatives, and not for
> identifier lookup.
>> About "we've decided that they don't get to use this character because it is inconvenient. ": Georg was saying "In Germany it is mandatory to maintain the small sharp s in uppercase names in official documents like IDs, passports or tax declarations etc.". In Japan it is mandatory to use Kanji characters for such documents which are not part of the Unicode repertoire, or can only be regarded as variants of Unicode code points. That does not prevent the use of Unicode code points in other areas - e.g. domain names. To put it differently: there are many characters that are not convenient for IDNs, and which hence are disallowed. If we have the requirement "IDNs need to be able to represent IDs of users", we open a big can of worms.
> I think this argument is indeed a problem if used as an absolute requirement.
> However, it is not useless as an additional data point. Arguing that
> the Germans shouldn't get something because the Japanese can't doesn't
> get us anywhere, and the same for the other way round.
agree, again that was not my intent. I am only wondering how to ensure
that we get the right input. Existing data from standards organizations
may not fit always well, as you mention. So currently we are just hoping
that we right people from various local communities follow this
discussion and give "representative" input. Probably there is no better
process than informally making communities aware of the discussion, a
encourage them to participate.
> Also, the history and circumstances of Kanji variants in names are much
> more fluid (you can even call them analog) than the German sharp s.
>> Felix (a German speaker living in Japan)
> Martin (a German-speaking Swiss living in Japan)
More information about the Idna-update