RE: FW: spoofing and IRIs (from Michel Suignard)

Here's what I don't understand:

An IRI is a sequence of Unicode characters. Is there not
already a well-defined way of converting a sequence of
Unicode characters to a visual display?

So how can there be a separate rule for converting
an IRI to a visual display which doesn't match the
Unicode rule?

Larry
 


-----Original Message-----
From: slim.amamou@gmail.com [mailto:slim.amamou@gmail.com] On Behalf Of Slim Amamou
Sent: Wednesday, March 03, 2010 3:46 AM
To: Larry Masinter
Subject: Re: FW: spoofing and IRIs (from Michel Suignard)

While security related issues can be pushed to UTR#36, I think the
"IRI BIDI structure" is of the utmost importance, even though I agree
that as it is framed now it could be replaced by a pointer to UTR#36.

I think this section should be rewritten to clearly define the visual
presentation structure of the IRI with regards to it's components :
the scheme, the authority, the delimiters,... And refer to UTR#36 only
for security considerations.

More specifically I propose to :
 * state that in a LTR context the components should be ordered (for
display) : scheme, authority, path, query, fragment and that the path
segments should be ordered segement1, segment2, segment3,
 * state that in a RTL context the components should be ordered (for
display) : fragment, query, path, authority, scheme and that the path
segments should be ordered segement3, segment2, segment1,

among other structure definitons.


On Tue, Mar 2, 2010 at 11:04 PM, Larry Masinter <LMM@acm.org> wrote:
>
> I like Michel’s proposal to replace section 4.2 Bidi IRI structure to a pointer to UTR#36, and encourage him to make this a concrete proposal (as an editor of both documents)


--
Slim Amamou | سليم عمامو
http://alixsys.com

Received on Wednesday, 3 March 2010 14:18:41 UTC