RE: Xml Schema profile

As I said in the article, I think the low occurrence of sGroups was due to the nature of the test cases.  I've gotten a
couple people following up with me regarding OAGiS.  Their version 8.x used sGroups extensively.  However, the version 9
that I used in the test did not.  But (and this was unfortunately put at the very end of the article), their usage of
global elements is intended to allow substitution via sGroups.  So while it did not occur in the base schemas, the
design reflected a need and desire to use them come implementation time.

I do think sGroups are a common extension point as well.  

The problem I see with sGroups is not that there is no need for it, but that tools do not always support it.  What I
have found with some consortia is that if the IDE supports it, then they think it is fine to use it.  And IDEs have
fairly decent support for sGroups.  The problem comes later when they try to use the schemas in a form generation tool
or a code generation tool.  It may be good news or bad news depending on the tool. 

I'd be interested in hearing what XBRL implementers and others say when they throw their schemas against a code gen
tool.  Are they happy with the tool support of sGroups?  When I was at HR-XML, we decided to create a wiki page
dedicated to tool issues because we were getting anecdotal information on some shortcomings and wanted to give folks a
heads up.  This is in part how we found that xsd:union was the least supported feature of schema.  But since HR-XML does
not use sGroups, I'd be interested in what others' are hearing from implementers.

Cheers,
Paul



W. Paul Kiel
XmlHelpline.com
"eXtensible Solutions"
work: 919-846-0224
cell: 919-449-8801
paul@xmlhelpline.com
Specializing in Xml, Xslt, web services, and data integration.


 

-----Original Message-----
From: Hugh Wallis [mailto:xmlschema@standarddimensions.com] 
Sent: Monday, September 25, 2006 4:59 AM
To: 'Boris Kolpackov'; Simon.Cox@csiro.au
Cc: paul@xmlhelpline.com; David_E3@VERIFONE.com; mike@saxonica.com; xmlschema-dev@w3c.org; lists@jeffrafter.com;
paul.downey@bt.com
Subject: RE: Xml Schema profile

Add XBRL (eXtensible Business Reporting Language) / http://www.xbrl.org to that list. It makes EXTENSIVE use of
substitution groups and has done so for a number of years. 

Thanks

Hugh

-----Original Message-----
From: xmlschema-dev-request@w3.org [mailto:xmlschema-dev-request@w3.org] On Behalf Of Boris Kolpackov
Sent: September 25, 2006 7:38 AM
To: Simon.Cox@csiro.au
Cc: paul@xmlhelpline.com; David_E3@VERIFONE.com; mike@saxonica.com; xmlschema-dev@w3c.org; lists@jeffrafter.com;
paul.downey@bt.com
Subject: Re: Xml Schema profile


Hi,

Simon.Cox@csiro.au <Simon.Cox@csiro.au> writes:

> Please do not labour under the impression that substitution groups are 
> a corner of the spec that rarely gets visited.

I was also somewhat surprised seeing this assumption. I did a quick search on our XML Schema repository which brought up
the following well-known schemas besides the opengeospatiol.org's ones (GML, etc.):

OVAL (Open Vulnerability and Assessment Language) / http://oval.mitre.org FpML (Financial products Markup Language) /
http://www.fpml.org COLLADA (COLLAborative Design Activity) / http://www.collada.org HR-XML / http://www.hr-xml.org


hth,
-boris


--
Boris Kolpackov
Code Synthesis Tools CC
http://www.codesynthesis.com
Open-Source, Cross-Platform C++ XML Data Binding

Received on Monday, 25 September 2006 12:16:33 UTC