NetKernel News Volume 2 Issue 11

January 7th 2011

What's new this week?

Catch up on last week's news here

Repository Updates

There are no updates this week. Steady as you are.

Support Cover

General support cover is now back to regular service. We're here to help if you need us.

NetKernel West 2011 - Registration is Open

We have finally nailed down all the details for the NetKernel West 2011 conference and I have pleasure to announce that registration is now open. You'll find descriptions of what's planned for the conference and bootcamp, together with details of venue, travel etc...

We've chosen The Armstrong Hotel as the venue this time. Its cool, its right in the heart of downtown Fort Collins, and its a stones throw from the biggest collection of micro-brew-pubs you've ever seen (not that we'd have any reason to need to visit those. cough.)

We've also managed to negotiate some amazing room rates (see conf page for details) but, just for you diligent newsletter readers, be aware the hotel is boutique so I'd recommend you make a reservation early to grab one.

As well as sweating the details on the conference organisation, one of my back-to-work tasks has been to knock up a conference booking and management application as a portlet for the NK services portal. (You'll see it when you register). Which was a double whammy since I also have it as a high-priority item to produce a tidy and documented version of the portal framework as an installable package. Watch this space.

The conf management system took a net 2-days to write. Half that time was refamiliarisation with, and enhancement of, the RDBMS schema. It's been tested as extensively as I was able in staging using Paypal's sandbox, but if for any reason you have any trouble with the user-facing side, just drop me a note and we'll sort it out manually ASAP using the admin tools. (As ever with these sorts of applications, the front-facing stuff is only the tip of the iceberg compared with the backend administrative tools that are needed.)

Looking forward to seeing many of you face to face in Colorado.

Transport Scope

It was a pretty quiet and peaceful break, but one small support item had me put down the sherry and mince pie for a minute or two.

Dan King at Orbis Technologies dropped me a short question for some advice on debugging a custom Transport he's developed. His problem was that his Transport used some configuration state and "for some reason" it wasn't finding it when it issued a request for it.

I thought I'd share the answer, since its something that, after the fact is relatively plain - but ahead of time could be less easy to anticipate.

Dan's spacial architecture looked something like this...

    <config>....some resource mappings for transport config stuff here...</config>
      <transport>The Misbehaving Transport</transport> ... other imports etc ...

So the transport's configuration state was being requested from, the apparently local, mappings in the mapper config. But they didn't resolve?

The thing to remember is that when you issue a request from a Transport, the space in which it is located is the root of the request stack (super stack). Requests from the transport therefore originate in the space in which its declared.

Ordinarily this is what you're used to, and typically a fulcrum space sits on the edge of your architecture - with external transport driven events causing requests to arrive into the ROC address space.

The subtlety in Dan's configuration was that the transport was declared inside the mapper (in the overlay space). So the transport considered that this space, was the top of the stack.

So the transport tries to issue a request for configuration resources and they are apparently declared in the mapper wrapping the space. But since the requests from the transport are originating inside the inner space they don't see the mapper at all - because the requests don't have a scope that goes above the inner space, its exactly as though the outer rootspace doesn't exist to them.

The moral here is that a transport originated request can never go up the superstack beyond the point where the transport is declared, since, to it, the space where the transport lives is the top. (There's logically nothing above it).

So that's the detail. It may be this knowledge is useful and the ability to keep transports isolated in inner spacial contexts could be appropriate in some, to be imagined, future pattern. But a general rule of thumb is, if you're declaring a regular transport, its probably best to put it in a rootspace.

NetKernel West 2011: Call for Papers

The NetKernel West 2011 conference will take place 13-14th April 2011 in Fort Collins, Colorado, USA. The conference will be preceded by a one-day NetKernel bootcamp on the 12th where you can get a fast track immersion in NK and ROC before the main event.

We want to make this an open opportunity for the NetKernel ROC community. We invite you to please let us know if you would like a slot for a talk/presentation/demo etc.

NetKernel West 2011
Location:Fort Collins, Colorado, USA
Conference:13-14th April 2011
NetKernel BootCamp:12th April 2011

Happy New Year, I've got a feeling this is gonna be a good one.


Please feel free to comment on the NetKernel Forum

Follow on Twitter:

@pjr1060 for day-to-day NK/ROC updates
@netkernel for announcements
@tab1060 for the hard-core stuff

To subscribe for news and alerts

Join the NetKernel Portal to get news, announcements and extra features.

NetKernel will ROC your world

Download now
NetKernel, ROC, Resource Oriented Computing are registered trademarks of 1060 Research

© 2008-2011, 1060 Research Limited