Re: [xsl] namespace values

Subject: Re: [xsl] namespace values
From: Peter Flynn <peter@xxxxxxxxxxx>
Date: Mon, 12 Mar 2001 22:18:57 +0000
On Mon, 12 Mar 2001, Mike wrote:

> At the risk of sounding like I am trolling, I will mention that very
> recently on the xml-dev list there was quite a bit of discussion about
> what "should" be returned if a namespace URI happens to be resolved as a
> real URL, even though there is no requirement that it actually be anything
> resolvable. For some reason, this time around, even though the issue has
> been raised many times before without incident, a mass spec writing circle
> jerk ensued (IMNSHO) and an overengineered solution to this, ahem, problem
> was born and named RDDL. If you read the spec at rddl.org, you'll notice
> that there really are no statements in it that justify its existence.

Oh dear :-) If I recall correctly, one of the original (long
since quashed) concepts behind a namespace was that if you wanted
to concoct a structure using (eg) lists-the-way-TEI-does, you
should be able to declare "tei" as a namespace and then reference
<tei:list> and <tei:item> and the relevant fragments of the DTD
in which those were declared would automagically be referenced
by traversing the implicit link provided by the URI in the
namespace declaration, at the end of which you would find a
modularised DTD in which the fragments could be located. Ditto
if you wanted to use <docbook:computeroutput>, or <html:img>.
But real-world DTDs aren't that granular, and a little digging
revealed that almost all such references would involve sending
out software agents to pursue nestings of element declarations
deeper than a politician's involvement in corruption :-) I think
your not-so-humble opinion above is dead accurate, Mike.

> Anyway, the fact that they have come up with this spec and seem eager to
> implement it means that you probably shouldn't lock yourself into using
> namespace URIs that are going to be difficult to resolve someday.

Using URLs that actually resolve right now could therefore be
dangerous of your document persists unedited and someone hauls
it up and expects to find something on the end of the hook.

///Peter
-- 
XSL: think like a tree, not like a chainsaw

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list


Current Thread