Re: [pkix] [apps-discuss] character repertoire for fragment identifiers, was: Fwd: FW: New Version Notification for draft-kerwin-file-scheme-13.txt

[Nico Williams <[e-mail]>]

On Mon, Jan 12, 2015 at 06:29:46PM +0900, "Martin J. Dürst" wrote:
> On 2015/01/12 07:10, Sean Leonard wrote:
> >***
> >I have empathy for what Sam/the W3C wants, since the HTML protocol slots
> >basically beg to be filled with Unicode strings like <a
> >href="http://zh.wikipedia.org/wiki/巴泰勒米·波岡達"> (instead of <a
> >href="http://zh.wikipedia.org/wiki/%E5%B7%B4%E6%B3%B0%E5%8B%92%E7%B1%B3%C2%B7%E6%B3%A2%E5%B2%A1%E9%81%94">).
> 
> Very very much so. The former is readable (although there are better
> examples than foreign names such as Barthélemy Boganda) to a
> significant part of the world's population; the later is gibberish
> for everybody.

The href attribute in HTML is one thing.  "_All_ URI/IRI slots" is
another.

I don't object to versioning protocols and data formats to change
specific URI slots into IRI slots.

Thus I have no objection to -say- the href attribute in HTML being made
into an IRI slot (with or without versioning of HTML -- that being an
issue for W3C, though if it were the IETF we'd want to do something
better than a flag day).

I object to any proposal that implies that all existing URI slots must
now [magically] be able to carry Unicode non-ASCII character data.

> >But maybe the more interoperable approach is to define a format and
> >mechanism (e.g., IRIs, or something like IRIs v2) to map /from ///the
> >Unicode-capable protocol slots, /to/ the well-standardized RFC 3986 URI
> >format.

We already have this: RFC 3987.

> With IRIs, that's essentially what we have. (of course I don't want
> to imply that the IRI spec cannot be improved upon)

Exactly.

Terminology is a problem here.  In conversation and these e-mail
threads, I really want to refer to IRIs as "URIs".  But in the context
of RFCs 3986 and 3987 it is critical to be careful to say "URI" in
reference to "slots" accepting US-ASCII only, and "IRI" in reference to
slots that notionally[*] accept Unicode.

It would be very convenient if we only needed to be so careful when
pairing "URI" and "IRI" with the word "slot".

[*] I say notionally because related UI elements generally can only
    handle an equivalent subset when used in non-Unicode locales.

Nico
--