RE: IDERF headaches

Yes, I just modified my examples generator to include a call to the
validating reader after each example is generated. Makes quite a difference.
I also picked up the fact that the QNames we declared were in the wrong
namespace in some cases.

The spec now has two build.bat files, any more complexity and I will start
using make. I rearrange the build file location though as it was causing
problems with some folks antivrius scanners.

		Phill

> -----Original Message-----
> From: Joseph Reagle [mailto:reagle@w3.org]
> Sent: Friday, July 19, 2002 4:41 PM
> To: Hallam-Baker, Phillip; Www-Xkms (E-mail)
> Subject: Re: IDERF headaches
> 
> 
> On Friday 19 July 2002 04:04 pm, Hallam-Baker, Phillip wrote:
> >     Turns out that our use of IDREF is not quite kosher.
> >     An IDREF has to reference an ID in the *same* document.
> >     So I have changed the offending IDREFs to string
> 
> I'm not sure what you are referring to, but if an attribute 
> value is making 
> a reference to an external document the correct type is 
> "anyURI" [1]. I've 
> never found a need for an IDREF in a schema. (The difference 
> between an 
> IDREF and an anyURI that point to a local ID is that the 
> anyURI will start 
> with the fragment identifier '#'.)
> 
> As an aside, in terms of debugging schema I learned that you can't be 
> confident until you've created an instance that excercises as 
> much as the 
> schema as possible and then run it through a validator (i.e., 
> xv or msv). 
> Sun also provides "xmlgen" that will create instances based 
> on a schema 
> which you can then use as a sanity check to look at, or run 
> through someone 
> else's validator!
> 
> http://www.w3.org/TR/xmlschema-2/#anyURI
>

Received on Friday, 19 July 2002 22:24:53 UTC