RE: [ciq] Creating schemas for other people's namespaces

ciq message

Subject: RE: [ciq] Creating schemas for other people's namespaces - what can you do and what can't you do?

Yes - W3C XInclude is another that publishes non-normative schemas. One reason is that you cannot create a schema for everything (some people may be surprised by this - but its actually quite easy to create well-formed XML that you cannot write a normative schema - for example scripting syntax - where you can only provide fragment syntax component construction rules).

In the case of XInclude and XLink - because the XIncludes and XLinks can be inserted at will into some other content - its pretty clear that you can create structures that a normative schema would cause to fail. Therefore the non-normative schema is illustrative of typical use cases.

Anyway - I'm really seeing that excessive use of namespaces for data elements leads to mydrid processing complexities. And therefore namespaces should be used minimalistically and only under extreme need.

One obvious use is to label pre-processing instructions within content - and XLink and XInclude are two excellent examples, where the namespaces denotes content to be excluded from the body of the main markup.

Unfortunately people use namespaces at whim and instead of being discouraged from this - are to the contrary - feel obliged to put namespaces on everything completely unnecessarily. In-line namespaces are especially troublesome.

We just have to live with this ugly namespace world - and this really is making XML much harder to do and implement.

Clearly with CIQ we want to publish clear normative instructions and use patterns to avoid these namespace quadmires for implementers.

Yet another reason to find robust direct and simple mechanisms for implementing business requirements.

-------- Original Message --------
Subject: [ciq] Creating schemas for other people's namespaces - what
can you do and what can't you do?
From: "Ram Kumar" <kumar.sydney@gmail.com>
Date: Tue, August 22, 2006 6:47 am
To: ciq@lists.oasis-open.org, "Max Voskob" <max.voskob@paradise.net.nz>

CIQ TC,

Have a look at this regarding the note from XBRL:

http://xsd.stylusstudio.com/2006Jun/post06001.htm

Ram