All, I've attached the latest and updated schema. There are two major changes. 1) Trivial - uses the latest Regular Expression for range values. 2) Big change: removal of jsdl:any and replacement with xsd:any##other. This is due to some big problems with the jsdl:any type. I'm going to put a summary below: a) It was noticed that the requirement for jsdl:any to always have an id tag was inconsistent with the rest of the document. We have stated for the rest of the document that an id tag was only needed when the multiplicity of an element (in the schema) is greater than 1. Thus forcing all extensions to the schema to have an id is incorrect. b) The definition that was there for jsdl:any did not match how I and (I think) most others expected it to work. Here is how I expected it to work: ... <jsdl:Resource> <jsdl:CPUTime operator="lessThanEqualTo">1000</jsdl:CPUTime> <globus:two_phase id="globusExtend">1</globus:two_phase> </jsdl:Resource> ... While in fact the document would need to look like the following: ... <jsdl:Resource> <jsdl:CPUTime operator="lessThanEqualTo">1000</jsdl:CPUTime> <jsdl:any id="globusExtend"> <globus:two_phase>1</globus:two_phase> </jsdl:any> </jsdl:Resource> ... This reduces the elegance of the document and makes it much harder to write. Michel, Nathalie, Willam, Donal and I have had a look at alternatives including using substitution groups and extending jsdl elements and all have some form of drawback which is either as bad if not worse than the fact we can't force people to use an id tag in their elements. As it has now been noticed that id tags don't always need to be there it seems much better to stick with the xsd:any##other tag in the schema and place into the spec document instructions on how extensions should be done and where id attributes should be placed. If someone could put the schema up on GridForge that would be great. Andreas - could you update the pseudo-schema in the document to use xsd:any##other rather than jsdl:any and we'll need to write the section in the spec to describe how to do extensions. steve.. PS if any XML guru knows how to achieve these rules in a schema please let us know. -- ------------------------------------------------------------------------ Dr A. Stephen McGough ------------------------------------------------------------------------ Research Associate, Imperial College London, Department of Computing, 180 Queen's Gate, London SW7 2BZ, UK tel: +44 (0)207-594-8310 fax: +44 (0)207-581-8024 ------------------------------------------------------------------------ Assistant Warden, West Wing, Beit Hall, Imperial College, Prince Consort Road, London, SW7 2BB tel: +44 (0)207-594-9910 ------------------------------------------------------------------------