Re: [I18nrp] Additional input needed for i18nRP BOF

[Nico Williams <[e-mail]>]

On Wed, Jun 06, 2018 at 10:38:30PM -0500, Adam Roach wrote:
> Thanks for taking a first stab at this. I think one of these criteria needs
> a bit of expansion:
> 
> On 6/6/18 10:30 PM, Nico Williams wrote:
> >  - does the protocol carry identifier strings (e.g., usernames,
> >    filenames, ...) or passwords?
> >
> >    Yes? -> I18NDIR review necessary; the protocol should specify some
> >            sort of stringprep, whether to normalize or allow
> >            form-insensitivity, etc..
> 
> This will also be the answer if you have any protocol that compares UTF-8
> strings, regardless of whether they are identifiers or passwords. For
> example, if one were to define a SIP header field whose value can contain
> characters greater than U+007F, and defined semantics for comparison of that
> value to another string, then such considerations would apply.

Sure.  Those would either be identifiers (if the protocol is open to
arbitrary headers) or protocol constants (whose exact encoding needs to
be specified without further I18N considerations).

Nico
--