Nameprep and ISO/Unicode (Re: [idn] NamePrep with Japanese hiragana and katakana)

[Harald Alvestrand <Harald@Alvestrand.no>]

At 15:04 14/12/2000 -0800, Hideyo Imazu wrote:

>That said, instead of introducing many new rules into the normalization
>stage, introducing the following entries into the mapping table is reasonable.
>
>KATAKANA-HIRAGANA VOICE SOUND MARK
>-> COMBINING KATAKANA-HIRAGANA VOICED SOUND MARK
>
>KATAKANA-HIRAGANA SEMI-VOICE SOUND MARK
>-> COMBINING KATAKANA-HIRAGANA SEMI-VOICED SOUND MARK

This comment is not intended as anything about the proposal above.... I 
simply do not have the knowledge to evaluate it, even though it "looks good".

But it serves well to illustrate a point.

There is a very real danger in nameprep, namely that much of what nameprep 
will be asked about involves repairing omissions or errors made in ISO 
JTC1/SC2/WG2 and the Unicode Consortium when constructing NFKC, NFC and the 
Case Folding UTR.

Some of it will be rooted in real differences between the rules underlying 
Unicode Consortium/ISO rules and nameprep requirements, and is Good.
Some of it will be "repairs" that we have to do, and is Necessary.
Some of it will be irrelevant to 99.99% of users, and is Mostly Harmless.
But some of it will be pushed by people who lost their battle in the 
Unicode Consortium and ISO, and are coming to the IETF to fight these 
battles once more, perhaps hoping for a different outcome.

And that is Bad.

Take care.

--
Harald Tveit Alvestrand, alvestrand@cisco.com
+47 41 44 29 94
Personal email: Harald@Alvestrand.no