Great Irish Government subsets of XML Standards

Smart government standards-leaders, implementers and technocrats should be alerted to the Irish Government's Reach Interoperability Guidelines (RIGs). These are profiles of the W3C recommendations: dumbed-down subsets aimed at being more useful and usable.

The key quote is Simply naming a set of W3C /OASIS / IETF recommendations will not produce interoperability.

Also Any existing interoperability standard that has not been through at least one profiling phase is unlikely to have been proven in the real world and thus should be treated with caution!

Published so far are:

• RIGs Terminology. (RIG 1).
• XML 1.0 Profile. (RIG 2).
• XML Namespaces Profile. (RIG 3).
• Schema Languages Recommendations. (RIG 4).
• Unicode and Internationalisation. (RIG 5).
• Draft - Versioning. (RIG 6).
• Draft - Interoperability Theory and Practice. (RIG 12).
• Draft - Reach Envelope. (RIG 100)
• Raser Agriculture: Message Service (RIG M005)
  
• Raser Revenue: Message Service (RIG M006)
  

...With a few dozen more on the way. A few technical papers are pretty interesting: Push Pull Overview talks about why to favour pull delivery of XML messages rather than push delivery. Actually, originating services push to the broker and receiving services pull from the broker. A little bird told me that this kind of dual queued messaging didn't necessarily fit in with the off-the-shelf capabilities of the current products lines of at least two of the big players,
but the customer is always right.

While I would quibble about a couple of details (who wouldn't?), the RIG project itself is a really positive step in the maturity of XML. And an exemplar for other goverments and their customers. Some things are Irish and non-applicable: the details in internationalization guidelines, for example.

Quibbles? Unicode is not a subset of ISO 10646; also, does the Namespaces profile contradict itself somewhat by advising Restrict namespace declarations to the root-element only but then there is no way to avoid namespace bleed other than to avoid namespaces on the root element of the Reach Envelope?; and the decision to avoid PIs knocks out using the profile with built-in stylesheets, such as CSS and XSLT, as it stands: but perhaps this is sensible.

A prose detective might find Sean McGrath's cadence throughout the RIGs, and they are a delight to read on style alone. For example, the word consolidation is used to describe what a profile does: so much better than simplification, which which is rightly an effect, not a cause.

And a double thumb's up for the mention of Schematron: when Schematron makes it through ISO then the RIG people will probably profile it too.

Also this was heartening to read, for the harmless drudges who make and review standards: Other would argue that the ponderous, deliberate work that goes into these (...large failed...) standards serves the vital role of acting as “idea pools” for smaller, more pragmatic standards. Evidence supporting the latter view can be found in the fact that the DSSSL standard (from ISO), inspired much of XSLT (from W3C) and SGML (again from ISO) directly begat XML (W3C).

Related link: http://sdec.reach.ie/rigs