NetKernel News Volume 2 Issue 34

June 24th 2011

What's new this week?

• Repository Updates
• Service HouseKeeping
• nk4um 1.1 release
• Metadata The Movie
• Meta-Metadata
• Euro-Bootcamp

Catch up on last week's news here

Repository Updates

The following updates are available in the NKEE and NKSE repositories...

• kernel 1.16.1
  • improve request debug for meta requests (non essential change)
• lang-dpml 1.15.1
  • updated documentation for built-ins
• layer0 1.64.1
  • fix to ensure validation cache persisted (big boot time improvement, thanks to Chris Cormack)
  • support for <meta> tag
  • fix leak in Boot Order Optimizer
• layer1 1.30.1
  • fix to dynamic import to stop spurious metadata failures being reported after a space change event
• module-standard 1.47.1
  • only single import warning message per issue during startup
  • support for <meta> tag
• nkse-doc-content 1.30.1
  • added documentation for <meta> tag support
• nkse-docs 1.15.1
  • improved endpoint datasheets (now show name/description/id from metadata, and also displays if id is autogenerated)
• nkse-search 1.13.1
  • Added active:textSearchEscape to escape Lucene operators in a query (thanks Chris again)
• system-core 0.25.1
  • support for <meta> tag - new userMetaAggregator accessor
  • late validationCache save call to catch post commission validations (performance improvement)
  • fix to prevent memory leak when ext-system is redeployed

Service HouseKeeping

You may notice that at the top of the newsletters there's now a link to an index page which provides a brief summary of all the news items in the current volume. There's a master index of volumes here too.

The index compliments the existing full text search - found in the top right of every page, a standard feature implemented in the wink wiki.

Its very simple to implement search using the text:search tools, based on the Lucene engine. Install wink to see how its done. Or alternatively take a look at nk4um since the wink wiki's search implementation has been contributed to the nk4um project. The general design principal is to implement the search index externally to the primary data persistence layer. The index is updated whenever a new item is posted or deleted. To seed the system there is also a master re-index capability.

The 1060 Research NetKernel forum is now using the latest nk4um build with the new contributed fulltext search (see next item)

nk4um 1.1 release

Chris Cormack has released a new build of the nk4um application. His change log is here.

Notable features are:

• Implementation of full text search.
• Pagination of topic views to make first page views faster if they're not cached.
• Pagination of search hit pages.

As I said, we're now running this in production and its smooth and fast.

Metadata The Movie

Following last week's resource oriented perspective on metadata, here's something I dug out of the archives for your amusement.

Its a short animated movie that John Erickson and I made in 2000 as a way to try to convey the intangible concepts behind metadata.

Maybe you'll be able to spot that the thing that interested me was not the surface details of the applications and services that are depicted. It was that we'd realised that classical software technology was not suitable.

Believe me, we were trying - we were researching electronic contracts, internet billing and payment aggregation, digital object identity and resolution etc etc (some of the screen shots in the background were real prototype systems). These were large scale highly ambitious - internet infrastructure scale.

As you know my background was in Physics, so the more I observed the mismatch between my ambition/expectation and the economic reality of what a development team could deliver, the more I felt like it was this economic failure of software that was actually the real research challenge.

So we progressively stepped away from the details of any one vertical domain problem. We (Russell Perry, Royston Sellman, Tony Butterfield and I) were working on a general XML resource model and processing system - DEXTER (Declarative XML Transform Engine).

It enabled us to take the first steps toward, and explore, a general resource oriented philosophy for software and systems architecture. For example, it implemented the active URI and had basic implementations of language runtimes (including a first generation DPML). Fundamentally, it allowed us to define and capture the economic properties of the Web and bring these down to solve arbitrary software problems.

In turn, Dexter evolved from XML, to embrace arbitrary resource models and, with a true-microkernel, became the general NetKernel platform. Armed with the insight that resources are the scale-invariant fabric of information systems, Tony and I founded 1060 Research and could fully articulate "Resource Oriented Computing".

Hopefully, after reading last weeks essay "On Metadata" you'll recognise that the story has gone full circle. I hope it showed that, in a self-consistent way, ROC offers a very compelling perspective for defining what metadata is. Which, if you keep on going round the circle, strongly suggests to me that if you're in the business of creating metadata-based solutions (who isn't?) then ROC is the correct foundation for them.

But then you would expect a circular argument - since it was in trying to economically implement large scale hardcore information systems that we realised we had to step away from code and understand the resource oriented domain!

Anyway, before we disappear up our own exhaust pipes, grab some popcorn and look out for "The man with two noses". Incidentally the brilliant animation and voice-over is by Alan Snow.

Meta-Metadata

Just as with the story behind the movie above. There is a picture behind the picture of metadata and ROC I painted last week.

Whilst ROC provides an elegant foundation in which to understand metadata. The point I'm really trying to drive home is that resource oriented computing provides a simplifying foundation to any information problem - metadata being a particularly intractable and intangible case in point.

The bigger picture is that embracing an abstract view of information resources changes the game.

And by playing a different game, you change the economics. Which, after all, is always the limiting factor on any solution.

Euro-Bootcamp

The Bootcamp that preceded the NK West conference was well attended and proved to be an effective way for many people to get their feet wet with NetKernel and ROC. We also know that due to the travel logistics many people outside of the US couldn't make the conference. So we're wondering if there'd be interest in us doing a one or two day bootcamp over on the East-side of the Atlantic?

This is a tentative idea at the moment and we're trying to gauge interest. But so far we have received a good response and its rapidly moving from speculative to being a commitment. Its still likely to be in the October time frame and somewhere easily accessible in Europe (most likely Brussels).

Drop me a note if this interests you and, all things being equal, we'll arrange things accordingly.