What's new this week?  We're in the process of coordinating a set of updates to some of the low-level infrastructure pieces - these aren't fixes but enhancements, like a bootloader that can be updated from apposite, discoverable module factories etc.  We want to ship these pieces as common to both NKEE and NKSE and get them out sooner rather than later before people are committed to production.

Since we want to have an updateable bootloader, we have to manually replace the current fixed bootloader, and we can most cleanly do that by shipping a new distro image. So watch this space for a NKSE 4.0.1 distribution update next week  (this will also incorporate all the post launch updates too).

Repository News:

XUnit - was updated with assertion coverage for <assert><int/></assert> that now is internally type neutral and will assert on int, long and BigInteger.  Ditto we added assert <float/> which does the same for float, double.

xml-saxon - a strong community effort has resulted in a complete update of this library to use the latest Saxon 9.2.  Each endpoint is now internally implemented using the Saxon S9API - which leads to a lot of internal refinement and also makes possible new streaming use cases and provides access to Michael Kaye's latest efficient document object model XdmNode.

The existing active:xslt2 and active:xquery services are interface compatible with the last generation.  They now also support an optional typemap argument that allows POJO arguments to be provided and if required transrepted before being presented the the procesor internals.

In addition a new accessor active:xpath2 provides a means to perform very rich POJO extraction operations on XML resources.  So for example it can be used to make a boolean assertion about the tree or existence of attributes,  can be used to extract string values from elements or attributes, can be used to extract sub-tree's from the doc.  Given that it is xpath2-based it also provides full access to the results of xpath2 functions (namespaces etc etc).

Acknowledgements go to Chris Cormack and Menzo Windhouwer - the community maintainers of xml-saxon - and of course to Michael Kaye.

ROC News:

Stung by the scathing public humiliation of being exposed for not blogging last week,  Tony has launched into a first discussion of the nature of ROC computational caching as a generalization of memoization. (Or maybe he was responding to a "NEW Blog-Entry+subject@Caching" instruction from me)

http://durablescope.blogspot.com/2009/11/caching-how-and-why-it-works.html

I have been mostly busy making tea this week (cf blog).

The 1060 Team