W3C home > Mailing lists > Public > www-svg@w3.org > June 2003

Re: xlink

From: Thomas DeWeese <Thomas.DeWeese@Kodak.com>
Date: Fri, 27 Jun 2003 08:05:43 -0400
Message-ID: <3EFC3317.1000708@Kodak.com>
To: Jim Ley <jim@jibbering.com>
CC: www-svg@w3.org

Jim Ley wrote:

>"Thomas DeWeese" <Thomas.DeWeese@Kodak.com> wrote in message
>>    I'm not sure I agree with Dean, because in the erratum it specifies:
>><!ENTITY % xlinkRefAttrs "xmlns:xlink CDATA #FIXED
>>    Since xlinkRefAttrs is present on all the relevant elements it sets the xmlns:xlink correctly.  So if the document specifies the SVG DTD then it does not need to provide xmlns:xlink, if it does not specify the SVG DTD then it must provide all the appropriate namespace specifications (SVG's for example).
>However SVG viewers are not required to have a Validating parser and therefore read the DTD to get at these #FIXED attributes. (although a non-validating one could do so)   With #FIXED the document can be different depending on if the DTD is read.  So the above is only true if a validating parser is used.
    True but isn't this equivilent to SVG saying that the 'user agent 
stylesheet' must set
overflow to hidden on elements that define viewports?  Similar to 
validation an SVG
user agent doesn't need to support CSS but it still needs to make the 
default for
overflow 'hidden'. I think if the document says to use the official SVG 
useragent is obliged to define the default namespace to SVG and the 
xlink prefix to
Xlink, even if it doesn't read the DTD, this is what the official DTD 
says the info set
should be (In fact all the 'real' XML parsers I am aware of _do_ read 
the DTD, even when not doing validation, but I don't know the XML specs 
well enough to know if this is required
or not).
Received on Friday, 27 June 2003 08:07:32 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 8 March 2017 09:46:56 UTC