RE: Agreement on IRI "processing spec" moving to W3C

The original motivation for taking up this work in the IETF in the first place was to make sure there was a common specification for what IRIs are, as well as how to process them, that allows, explains, or at least gives caveats about, how to copy/paste to/from HTML into/out-of other contexts that use internationalized resource identifiers/locators, including XML-based contexts that use LEIRIs and other systems that wish to restrict themselves to a more narrow range.

It is understandable that there is a desire to specify, for HTML processors, a more liberal parsing algorithm which does pre-processing ("of a URL input string to a URL"), but it is also likely that there are other applications other than HTML which might wish to use the same input methods; at least that was the reasoning when IRI WG was chartered in the IETF.

Moving the specification of this processing from one parent organization to another (among WHATWG, W3C, IETF) just seems to me to be moving the shells around, since the same folks (implementors of URL parsers and processors and generators) need to be (and seem to be) involved in both groups.

HTML WG agenda:
http://lists.w3.org/Archives/Public/public-html-wg-announce/2011OctDec/0017.html

http://lists.w3.org/Archives/Public/public-html/2011Nov/0194.html


It seems like there are editorial tasks and specification tasks, that the specification tasks are more important, and the questions remaining are ones that could more readily be determined by a more extensive test suite/examples.

At the IETF meeting, the important resolution is to make sure the IRI specification and processing rules are actually consistent with the IDN rules, so that the rules for IRIs that contain internationalized domain names match the rules for internationalized domain names (or at least that we can explain the differences).

Larry

-----Original Message-----
From: Anne van Kesteren [mailto:annevk@opera.com] 
Sent: Thursday, November 24, 2011 3:39 AM
To: Frank Ellermann; Chris Weber
Cc: Martin J. Dürst; public-iri@w3.org
Subject: Re: Agreement on IRI "processing spec" moving to W3C

On Thu, 24 Nov 2011 08:36:32 +0100, Chris Weber <chris@lookout.net> wrote:
> There happen to be one, maybe two, attributes in HTML which allow for 
> a space-separated list of URIs.  That's the only reason this 
> statement, and the SP explicitly, was included in the pre-processing steps.

This is backwards, for what it's worth. The URL specification has no need to think about delimiters. It only needs to define the mapping of a URL input string to a URL. That HTML sometimes whitespace-separates URL input strings should not be of any concern.


--
Anne van Kesteren
http://annevankesteren.nl/

Received on Tuesday, 29 November 2011 18:59:40 UTC