Re: [xml-dev] xml:base and fragments

["C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com>]

> On May 4, 2017, at 2:09 PM, John P. McCaskey <mailbox@johnmccaskey.com> wrote:
> 
> This seems the opposite of Eliot’s initial reaction (and mine and I thought Simon’s but maybe not).
> 
> Just to be sure then, in my example, a file named example.xml.
> 
>     <div 
> xml:base="http://www.dictionary.com/a.html";>
>     
>   <p>
>         <ref target="#apple">Apple</ref>
>       </p>
>     </div>
> 
> The new Kay-informed (and already TEI majority) consensus is that @target points to a node in example.xml, not to one inside the file at a.html.

I don’t think that’s the majority view in the TEI discussion.  Or possibly I’m not in the majority over there after all.

The view that I think is entailed by RFC 3986 is that @target is a relative URI whose absolute form is http://www.dictionary.com/a.html#apple and that 3986 defines this resource as being located in the resource example.xml.  

I think the natural way to interpret this is that through an otherwise unexplained process, both the URI 

    http://www.dictionary.com/a.html#apple

and the URI

    whatever-goes-here/example.xml#apple

denote the same thing, which can be retrieved either by scrolling to the appropriate location in example.com (which is already loaded) or retrieving a.html from www.dictionary.com.  RFC 3986 recommends the former; the document() function in XSLT prescribes the latter.

I do not see any consensus in the TEI discussion for the proposition that the target attribute in the example does not point to http://www.dictionary.com/a.html#apple


********************************************
C. M. Sperberg-McQueen
Black Mesa Technologies LLC
cmsmcq@blackmesatech.com
http://www.blackmesatech.com
********************************************