I'm not a fan of the "multibyte" term at all and would love to see it
removed completely. We should, of course, do that through WG14 first
though since that is where the term originates. That would probably be a
fair amount of work for not a whole lot of gain though. And certainly
not something we should try to do via an NB comment!

If an NB comment is filed, please do strive to provide wording intended
to resolve it.

Tom.

On 8/22/22 10:59 AM, Corentin Jabot via SG16 wrote:
> The definition is wrong in a vacuum, but it's only ever used in the
> wording of C conversion functions, iostream, etc which have bigger issues
> (and that's all not super recent).
> I'd argue that the definition does not have any reason to be
> introduced in [intro] and should be moved in [character.seq].
> I'm neutral on whether the definition should be changed in the
> contexts in which it is used, given all these places will need more work.
> We could try to remove the use in [time.format] though.
>
>
> On Mon, Aug 22, 2022 at 4:46 PM Steve Downey via SG16
> <sg16_at_[hidden]> wrote:
>
> https://eel.is/c++draft/intro.defs#defns.multibyte
> " sequence of one or more bytes representing the code unit
> sequence for an encoded character of the execution character set"
>
> We're reviewing the draft internally and discussing possible NB
> comments. It wasn't clear to one of the reviewers, nor to me, that
> 'of the execution character set' is a necessary part of the
> definition, here. Additionally, in trying to explain what we mean,
> it looks like we don't clearly state what we mean by 'execution
> character set', although I think we have a clear understanding
> that it's how the library interprets text and is controlled by
> locale. Should we actually define it as part of defns?
>
> I thought it worthwhile to bring this up now before bringing it up
> in an NB comment, though.
>
> Thoughts?
> --
> SG16 mailing list
> SG16_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/sg16
>
>