- From: iri issue tracker <trac+iri@trac.tools.ietf.org>
- Date: Sun, 11 Mar 2012 08:45:56 -0000
- To: draft-ietf-iri-3987bis@tools.ietf.org, duerst@it.aoyama.ac.jp
- Cc: public-iri@w3.org
#117: conformance requirements in bidi document -- do they belong?
Comment (by duerst@…):
Larry writes:
> The document says "Bidirectional IRIs MUST be rendered by using the
Unicode Bidi IRIs MUST be visually ordered by the Unicode Bidirectional
Bidirectional Algorithm [UNIV6], [UNI9]."
This isn't the relevant bit of text, because rendering by the Bidi
algorithm isn't the issue, it's what context or additional tweaks should
be allowed (or not)
> There was significant discussion on the working group mailing list that
this advice should not be normative.
It's not about being normative or not. It's about whether another way of
rendering, or several ways of rendering, should be allowed.
> What is the consequence of not following this bit of advice anyway?
The consequence is that bidi IRIs get garbled; components get moved
around, and users get confused. Maybe to some extent unavoidable, but
definitely not a good thing.
--
------------------------+---------------------------------------
Reporter: masinter@… | Owner: draft-ietf-iri-3987bis@…
Type: defect | Status: new
Priority: major | Milestone:
Component: 3987bis | Version:
Severity: - | Resolution:
Keywords: |
------------------------+---------------------------------------
Ticket URL: <http://trac.tools.ietf.org/wg/iri/trac/ticket/117#comment:1>
iri <http://tools.ietf.org/wg/iri/>
Received on Sunday, 11 March 2012 08:46:36 UTC