W3C home > Mailing lists > Public > public-html@w3.org > June 2008

Re: Placement of <base> & xml:base interaction

From: Ian Hickson <ian@hixie.ch>
Date: Fri, 6 Jun 2008 00:18:13 +0000 (UTC)
To: Henri Sivonen <hsivonen@iki.fi>
Cc: HTMLWG WG <public-html@w3.org>
Message-ID: <Pine.LNX.4.62.0806060015090.8559@hixie.dreamhostps.com>

On Tue, 9 Oct 2007, Henri Sivonen wrote:
> The spec contains this sentence regarding <base>: "A base element, if it 
> has an href attribute, must come before any other elements in the tree 
> that have attributes with URIs."
> Is that sentence intended to be merely a statement of the consequence of 
> the content models of <html> and <head>? If so, it would be nice have 
> the sentence start with something that labels it as a restatement of 
> something else.

Apparently I've removed the requirement from the content model part, so 
that it is now a separate and non-redundant requirement.

> OTOH, if the sentence itself is meant to have the consequence that 
> xml:base must not appear on <html>, <head> or <meta charset>, it would 
> be good to state that consequence explicitly.

This is not intended.

> As for processing, exempting <base href> from xml:base processing and 
> saying that <base href> establishes the base RFC 3986 base URI for the 
> purposes of xml:base means that xml:base on <html>, <head> or <meta 
> charset> is not disrupted by <base href>. Therefore, the consequence I 
> speculated about in the previous paragraph seems excessive. (Even in the 
> case of streaming processing, it would be feasible to defer xml:base 
> computations until a conforming <base> has been seen.)

I assume this is thus not an issue.

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Friday, 6 June 2008 00:18:53 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:34 UTC