W3C home > Mailing lists > Public > whatwg@whatwg.org > February 2005

[whatwg] Use IRI instead of URI

From: Ian Hickson <ian@hixie.ch>
Date: Mon, 7 Feb 2005 16:59:19 +0000 (UTC)
Message-ID: <Pine.LNX.4.61.0502071655070.24987@dhalsim.dreamhost.com>
On Mon, 7 Feb 2005, Anne van Kesteren wrote:
> > 
> > Actually, I just went through the entire spec and made sure that every 
> > reference to URI is correct and every reference to IRI is correct. Let 
> > me know if you spot any cases that say URI when they should said IRI. 
> > I think they are all now correct, though.
> 
> 2.5. Extensions to existing attributes
> 
> Under 'maxlength': "Valid URIs and e-mail addresses in particular can 
> often be surprisingly long."
> 
> I guess that should be IRI since it refers to the value |type="uri"| can 
> have.

I left that one because the term "URI" is more likely to be understood 
than "IRI", and that sentence is aimed at explaining a conformance 
criteria to authors. (Note that that sentence is not itself a conformance 
criteria but a statement.)


> 2.12. The datalist element and the list attribute
> 
> It could be that the UA has listed some IRIs as well though I may be
> nitpicking here.

Again, the occurances here are either in explanatory text or in examples, 
which are themselves non-normative by definition.


> 2.16. Extensions to the form element
> 
> Not sure about this. If it is an empty string, should it really be
> relative to the base URI? I can remember a discussion in Bugzilla about
> same-document references where there still isn't a conclusion about
> this. (Among other things with regard to this.)
> 
> Or does this update the RFC in question?

The Bugzilla discussion to which you refer was regarding fragment 
identifiers; there are none in this case.


> 6.1. Filling select elements
> 
> Why "URI or IRI"? IRI covers both. Same for other sections where this is 
> used. (See for another example 6.2.)

Again, because the term "URI" is much more likely to be understood than 
"IRI". Mentioning both means that the conformance criteria is exact while 
still making the intent clear.


> A. XHTML module definition
> 
> A lot of attribute can only contain URIs where it would make sense to
> allow IRIs.

AIUI, we are restricted by XHTML M12N here.


> > > Also, I think WF2 should refer to RFC3986 instead of RFC1738 since 
> > > the former updates the latter. RFC3986 is an internet standard and 
> > > obsoletes various other documents as well.
> > 
> > I'd updated all the relevant reference, I think. Did you see some I 
> > missed?
> 
> I was wondering why you were pointing to RFC1738 as well as pointing to 
> RFC3986. However, as you explained on IRC RFC3986 only updates the 
> specification it does not obsolete it.

Indeed.

HTH,
-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Monday, 7 February 2005 08:59:19 UTC

This archive was generated by hypermail 2.3.1 : Monday, 13 April 2015 23:08:21 UTC