Re: [I18nrp] draft-faltstrom-unicode11-04.txt

[Asmus Freytag <[e-mail]>]

On 10/11/2018 2:17 PM, Nico Williams wrote:
> Future codepoint assignments could all begin life as unstable and gain
> stability with experience.

 From Unicode's perspective, the code point assignments are not unstable 
(what is encoded and where) is not changed. What changes occasionally is 
the understanding of character properties that relate to how a code 
point is used (including in certain software).

While identifiers are a critical use of code points, they are by far not 
the most important one. Therefore, usage information (character 
properties) cannot be arbitrarily frozen.

In some cases, property data does have stability, because reliable 
information is available at design time (case, for example). However, 
whether something is better seen as a combining mark or a punctuation is 
subject to change - unfortunately that is a property picked up by IDNA2008.

Especially where this happens with characters for obscure scripts or 
languages, there's not a lot of benefit of trying to address that with 
an exception.

The real issue is that obscure scripts and languages (i.e. dead ones) 
have no business being supported in a public zone.

That's a different kettle of fish.

A./