I've just done some testing, and it seems that an old bug using =
xs:redefine to restrict the occurrence range of a group from 1-to-many =
to exactly 2 is still plaguing XSV. (See =
http://lists.w3.org/Archives/Public/xmlschema-dev/2005Feb/0041.html, =
noting that the message is different.)

in http://www.ruleml.org/temp/xsv-bug/restriction.xsd:
Atom :: slot*, ( arg+, slot )?

in http://www.ruleml.org/temp/xsv-bug/restriction2.xsd:
Atom :: slot*, ( arg, arg, slot )?

XSV reports slot/slot non-determinism (pasted below) when validating =
http://www.ruleml.org/temp/xsv-bug/restriction.xml (which points to =
restriction2.xsd). If this file is changed to point to restriction.xsd, =
all is well.

Is this a known issue (or am I missing something)?

Dave

***

Detected during instance validation

http://www.ruleml.org/temp/xsv-bug/restriction.xsd:22:2: Invalid: =
non-deterministic content model for type Atom.type: =
{http://www.ruleml.org/0.91/xsd}:slot/{http://www.ruleml.org/0.91/xsd}:sl=
ot