ISO member bodies vote to approve committee draft of new version of Schematron!

By Rick Jelliffe
July 16, 2010

Opening the champagne here in chilly blue-skied Sydney: the Committee Draft of the new version of ISO Schematron standard which I edit has been approved by a very good majority: 20 national bodies voted to approve as submitted, 9 national bodies abstained, 1 voted approve with a comment (an improvement on bidirectionality) and 1 voted no with comment (but indicated they would change their vote if their issues were dealt with adequately.)

So the next step will be for WG1 to prepare a Disposition of Comments (to directly address each of the comments from National Bodies), prepare a revised text (even though the vote was overwhelmingly to approve as submitted, an ISO committee always tries to satisfy as many comments as possible) and make its recommendation for further processing. I expect it will head down the sausage machine.

So this is a pretty satisfying day for me: the revision has been a long time coming. I'd like to thank all the reviewers and participants in the process. By the time the standard is actually out, I hope most users of the skeleton XSLT implementation will have changed over: the new draft is implemented in the current skeleton code. The update includes better support for XLST2, a better extension mechanism, properties and multi-document validation (each pattern can have different document).

(For people with ISO SC34 passwords, the ballot results are here.)

What about the country that voted 'no with comments'? I don't think I can say too much about details, due to the requirements for discretion, but their comments seem very good and sensible, with no surprises: the comments come from industrial experience in using Schematron by experts who know the standard well, so I really welcome the comments and take them seriously. I favour minimal standards with plenty of wriggle room for young standards, and it does not surprise me that the current text does not go quite far enough for other experts and that Schematron is mature enough that more guidance is appropriate. The way things work now is that as editor I talk around and prepare an initial disposition of comments, which are suggestions for how to satisfy the comments, which WG1 then uses as one of its inputs for figuring out the final text.

In other Schematron news, I am midway through moving my open source 'skeleton' XSLT implementation over to Google Code. It will include Schematron, the various input and output converters like the embedded Schematron extractors for RELAX NG and XSD, the Ant Task for Schematron, and the XML Schemas to Schematron converter XSD2SCH. Paul Herman has indicated he will provide his test suite for XSD2SCH, and I have managed to get more explicit statements about licensing (and to change over the license to the MIT license) from most of the contributors to the source code (I don't think these are necessary, but it can simplify enquiries from Eclipse, Apache and people wanting to incorporate Schematron into commercial products.)


You might also be interested in:

News Topics

Recommended for You

Got a Question?