Re: Importing xmldsig-core stuff...

On Tuesday 25 May 2004 12:56, Stephen Farrell wrote:
> The RFC editor pointed out that their schema checker objects to
> the schemaLocation value above since they don't have a dsig
> core schema at that relative location.

That sounds approriate.


> The schemaLocation attribute is however just a hint [2], so the
> schema passes using the checker on the w3c site.

Yes, though if you do use it, the Schema spec says [b] "When a 
schemaLocation is present, it must contain a single URI reference which the 
schema author warrants will resolve to a serialization of a ·schema 
document· containing the component(s) in the <import>ed namespace referred 
to elsewhere in the containing schema document." 

[b] http://www.w3.org/TR/xmlschema-1/#composition-schemaImport

> So, the question:-
>
> - is it ok as is?

A non-working hint doesn't seem appropriate to me, why would you use it?

> - should I change the value of schemaLocation to use an
>    absolute URL, in particular [3]?
>
> I assume the latter is better.

Yep, the only way it wouldn't is if somehow one interpreted the URL as a 
secure string/identifier. For example, there's some program that will only 
use schemas that when they import from dsig, import that specific string, 
and not the URL. But I'm not aware of any such processing, and the W3C 
never indicated "../W3C/xmldsig-core-schema.xsd" is a blessed 
string/identifier -- unlike other algorithm identifier URIs where we use an 
absolute URI.

Received on Thursday, 27 May 2004 15:17:30 UTC