From: Richard Wordingham (firstname.lastname@example.org)
Date: Wed May 10 2006 - 15:53:04 CDT
Philippe Verdy wrote on Wednesday, May 10, 2006 at 8:40 AM
> From: "Richard Wordingham" <email@example.com>
>> Philippe Verdy wrote on Tuesday, May 09, 2006 3:48 PM
>>> the stability of normalized forms, and so should better not apply any
>>> normalization, to keep the texts intact (this is a conforming behavior,
>> Please give an example of how normalising text with an undefined
>> can corrupt the text.
> I did not use the terms "corrupt the text", only you did, assuming that
> normalization instability means a text corruption...
Neither of us meant what the other thought. I thought you meant one should
avoid normalising such texts so as not to corrupt them, but I couldn't see
how correct normalisation would corrupt them. The point about normalisation
stabilisation is that once Unicode-compliant text is in NFC, or NFD, or
NFKC, or NFKD, it should remain in that form forever even as the Unicode
standard develops. This matters if inputting a text to a non-Unicode
I would say that if non-compliant text (because it contains an undefined
codepoint) needs to be in NFC, convert it or reject it. The problem is that
it will need reconverting if it becomes less non-compliant with Unicode but
no longer NFC.
This archive was generated by hypermail 2.1.5 : Wed May 10 2006 - 15:55:04 CDT