On Sun, 21 Jun 2020 at 13:45, Jens Maurer <Jens.Maurer_at_[hidden]> wrote:

> On 20/06/2020 15.16, Corentin Jabot wrote:
> > There is a simple model where everything is done in the source character
> set directly in phase 1 to 5,
> > We don't need to describe what characters are in that set, because there
> is no process by which phase 1-5 creates extra characters.
> > And we do not need to talk about ucns before phase 4+ and 5.
>
> We want to ensure macro names (identifiers) are NFC.
> Discussing this only makes sense for characters in Unicode.
> Having an island of Unicode-transformed macro names in phase 4
> (but keeping the rest in the original source encoding) seems odd.
>

Agreed!


> Jens
>