Fellow members of UBL SBSC,

I would like to propose to the SBSC that we as a group consider a different approach towards specifying the UBL 2 SBS than was done for the UBL 1 SBS.

As you recall, we specified UBL 1 SBS exclusively through XPath files that catalogued the information items in the subset.

In the development of UBL 2 there has been feedback regarding the perceived necessity, let alone convenience, to some of having schema expressions of the subset. This would be a set of schemas for the same namespace but specifying a lesser set of optional constructs, just as was specified by the XPath files, but doing so using XSD.

I've posted some thoughts of mine regarding UBL customization approaches here:

http://lists.oasis-open.org/archives/ubl/200606/msg00109.html

Please let me know your thoughts about treating UBL 2 SBS as a UBL customization effort such that we produce an SBS set of schemata, as described in chapter 8 of that paper referenced above. As I've indicated in that paper, I propose using XPath files to conclusively prove through exhaustive enumeration that each and every possible information item and cardinality of the SBS schemata is an information item allowed in UBL. I don't see the SBS needing any extension elements, only needing to elide optional elements.

Because of the push for code lists and other UBL 2 issues (and writing my UBL tutorial) I have yet to write the code utilizing XPath files to prove the exhaustive enumeration, but this is high on my "to do" list once we get into UBL 2 review.

Nevertheless, I thought it important that we should start talking about UBL 2 SBS now in order to get ideas on the table for consideration. Hopefully for UBL 2 SBS we would be able to subset the common library expressions and each of the document types such that we have a single suite of XSD expressions of the normative subset of UBL 2 that is the Small Business Subset.

Please let me know what you think.

. . . . . . . . . . . . Ken