W3C home > Mailing lists > Public > public-html@w3.org > June 2007

Re: Do we need the restrictions on the <base> element?

From: Laurens Holst <lholst@students.cs.uu.nl>
Date: Fri, 08 Jun 2007 10:52:06 +0900
Message-ID: <4668B646.4000102@students.cs.uu.nl>
To: Boris Zbarsky <bzbarsky@MIT.EDU>
CC: "public-html@w3.org" <public-html@w3.org>
Laurens Holst schreef:
> Here I am suggesting that /automatically/ re-evaluating and updating 
> included resources when their base URI changes, e.g. when moving them 
> or changing the xml:base attribute, would NOT have to be done, while 
> still complying with standards (and possibly explicitly define this in 
> HTML5). You reply that you think it’s kind of odd behaviour, upon 
> which I agree, but say "well if that’s what it takes, it’d work for 
> me". Then I continue to say that UAs already do exactly this (maybe I 
> should add here again that Firefox does a different thing with <base> 
> than all other UAs), give <base> and xml:base as examples, saying 
> "Firefox also implements xml:base that way".

So what do you think about that behaviour?

I suggest that:

   1. HTML5 defines <base> as it’s used in Internet Explorer and Opera,
      which would mean that the baseURI property on the document object
      reflects the current value of the <base> element, if present.
   2. For all included resources, HTML5 defines explicitly that their
      address is not automatically re-evaluated when their base URI
      changes (either through <base> or xml:base). This would be
      consistent with the current implementation of <base> in Internet
      Explorer and Opera, and consistent with the current implementation
      of xml:base in Firefox. Quoting performance as reason to deviate
      from what would be most logical.

Although if we could find a way to make all included resources 
automatically update when their base URI is changed while still 
performing well, or if we could conclude that automatic updates would 
not really impact performance too negatively (after all, the DOM 
NodeList is also ‘live’), it would be even better to make point 2 define 
the opposite (quoting the well-performing method as an example).


Ushiko-san! Kimi wa doushite, Ushiko-san nan da!!
Laurens Holst, student, university of Utrecht, the Netherlands.
Website: www.grauw.nl. Backbase employee; www.backbase.com.

Received on Friday, 8 June 2007 01:52:27 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:22 UTC