HTML5/IRI: "resolve a URL"

[Julian Reschke]

>From <http://dev.w3.org/html5/spec/Overview.html#urls>:

"To resolve a URL to an absolute URL  relative to either another absolute URL or an element, the user agent must use the resolve a Web address  algorithm defined by the Web addresses specification. [WEBADDRESSES]"

WEBADDRESSES (<http://www.w3.org/html/wg/href/draft>), when defining this, has lots of references to the HTML "Document" object.

These aren't going to end up in IRIbis, right? So it appears that IRIbis will not be able to replace WEBADDRESSES as a reference completely. Does stuff need to move back from WEBADDRESSES into HTML5?


--
Jetzt kostenlos herunterladen: Internet Explorer 8 und Mozilla Firefox 3.5 -
sicherer, schneller und einfacher! http://portal.gmx.net/de/go/atbrowser

[Anne van Kesteren-2]

On Tue, 03 Nov 2009 14:57:48 -0800, Julian Reschke <julian.reschke@...>  
wrote:

>> From <http://dev.w3.org/html5/spec/Overview.html#urls>:
>
> "To resolve a URL to an absolute URL  relative to either another  
> absolute URL or an element, the user agent must use the resolve a Web  
> address  algorithm defined by the Web addresses specification.  
> [WEBADDRESSES]"
>
> WEBADDRESSES (<http://www.w3.org/html/wg/href/draft>), when defining  
> this, has lots of references to the HTML "Document" object.
>
> These aren't going to end up in IRIbis, right? So it appears that IRIbis  
> will not be able to replace WEBADDRESSES as a reference completely. Does  
> stuff need to move back from WEBADDRESSES into HTML5?

Yes. That should probably be sorted out when iri-bis is somewhat more  
stable.


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

["Martin J. Dürst"]

Hello Julian, Anne,

Thanks for pointing out what may turn out to be a backreference (or
"free variable") in the IRI spec. We will try to limit these to the
minimum. One other one that I already cursorily discussed with Ian is
"white space". HTML5 has a very specific definition for that which may
not be the same for other formats (e.g. CSS?).

Regards,   Martin.

On 2009/11/04 12:53, Anne van Kesteren wrote:

> On Tue, 03 Nov 2009 14:57:48 -0800, Julian Reschke
> <julian.reschke@...> wrote:
>>> From <http://dev.w3.org/html5/spec/Overview.html#urls>:
>>
>> "To resolve a URL to an absolute URL relative to either another
>> absolute URL or an element, the user agent must use the resolve a Web
>> address algorithm defined by the Web addresses specification.
>> [WEBADDRESSES]"
>>
>> WEBADDRESSES (<http://www.w3.org/html/wg/href/draft>), when defining
>> this, has lots of references to the HTML "Document" object.
>>
>> These aren't going to end up in IRIbis, right? So it appears that
>> IRIbis will not be able to replace WEBADDRESSES as a reference
>> completely. Does stuff need to move back from WEBADDRESSES into HTML5?
>
> Yes. That should probably be sorted out when iri-bis is somewhat more
> stable.
>
>

--
#-# Martin J. Dürst, Professor, Aoyama Gakuin University
#-# http://www.sw.it.aoyama.ac.jp   mailto:duerst@...