[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
Re: [xml-dev] Should information be encoded into identifiers?
- From: rjelliffe@allette.com.au
- To: "'xml-dev@lists.xml.org'" <xml-dev@lists.xml.org>
- Date: Sun, 7 Mar 2010 16:21:09 +1100
> On Sun, 2010-03-07 at 14:44 +1100, rjelliffe@allette.com.au wrote:
> The only reliable way to do this is not to inspect the URIs but to use
> HTTP HEAD and check the MIME media type of whatever you get back. Of
> course, a server could randomly send a different format each time...
> unless you own the server to which the URI refers, all bets are off.
> This decoupling (as you know, Rick, but many people reading this might
> not, which is why I need to say it), this decoupling is fundamental to
> Web architecture.
Yes indeed, good point . That is why I put in the clause "in particular in
closed systems where the IDs are generated automatically": on the WWW you
have no control of things, and you cannot know the representation's
notation from the URL, including the extension. However, the opposite it
true behind the scenes in your closed system, where you do have control
and knowledge.
I think XSLT, Schematron and XSD are rarely used on data retrieved from
public URLs however: they are more likely to be used on data you yourself
share some control of... (However, my second example was probably that
kind of access to the WWW, so I have brought it on myself.)
> It's also why XML NOTATION was broken as designed.
Ultimately, I suspect the breakage in XML NOTATION was that XML Schemas
did not bring out the connection between a notation and lexical types
inter alia. Though perhaps this is just on the grounds that all problems
can and should be traced to XSD :-) As an off-ramp for SGML to XML, I
think NOTATION was a reasonable thing to provide. As a thing in itself
nowadays, I agree it is not popular which is a reasonable (but not the
only) test of the value of some part of a mass (rather than niche)
technology.
> Currently, this isn't something you can do in XSLT - there's another
> thread about exactly that, of course.
So you are saying, "don't do that, it can break; instead do this, it
definitely won't work!" :-)
Cheers
Rick
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]