Jena 2 Upgrade

Just a quick note to say that I upgraded the FOAF-a-Matic Mark 2 to use Jena 2.0.0-beta1 and found the migration pretty painless. I’d read Shelley’s Jena week articles as preparation so had a good idea about what I needed to do.
For the most part it was just changing imports, and switching to use ModelFactory rather than explicitly creating ModelMem objects. I tweaked my I/O code to use Streams rather than Readers/Writers as described in the I/O Mini HowTo (definitely a must read), and reworked the code to set namespace prefix mapping following Shelley’s recipe.
Took me roughly an hour including reading through the HowTo and skimming the new Javadocs for those bits of the API I was using. And it all worked first time. Congrats to the Jena team.
One item that Shelley didn’t mention, was that the Selector interface has changed. It now has a few extra methods to implement, so if you’ve got custom Selector implementations you’ll need to check that out. In most cases however you can switch to using the new SimpleSelector class (rather than the older SelectorImpl) and get the same behaviour.
Oh, and change to RDFException so that it’s no longer a checked exception is another welcome one!
This is another item ticked off my list for moving Mark 2 into a second beta (long overdue). I’ve already provided a dev snapshot for Phil Wilson so let me know if you’re interested in seeing what I have before then.

FOAF Concerns, revisited and discussed

Shelley Powers has posted a list of concerns that she has with FOAF as currently defined and used:
Practical RDF: Danger, Will Robinson: Part 2
To summarise, these are:

  1. Inappropriate use of Myers Briggs property
  2. What does foaf:knows mean?
  3. Social implications of foaf:knows
  4. Publishing data about other people

For more details on each see Shelley’s post.
These set me to wondering about these concerns might be addressed. Here’s some thoughts aimed at promoting discussion only. I’m not suggesting I’ve thought of all the angles. Take ’em as a starter for ten.

Read More »

BBC Archives Going Online?

Excited to see this piece:
Dyke to open up BBC archive
It’d be great to be able to access archives of the BBC’s output online, especially if its reasonable quality. Who needs a Tivo?!
I hread rumours that the BBC’s archives are in a hopeless mess, so I wonder how long it’ll take them to get it all online, and how much it’ll cost.
Raises some interesting questions about licence fees as well, as presumably the content will be available to anyone on the web. Personally I don’t mind sponsoring the BBC to do this, but I’m not sure everyone else will.
All sorts of interesting possibilities spring up if the material is freely available, including collaborative annotation, categorisation, reviews, etc. Lets hope they make it available within as clue-ful an environment as possible.

Dynamic Languages and Linking

Ted Leung points to some interesting comments from Jon Udell and Don Box about Dynamic languages, virtual machines, information clients, and more.
This is related to some musings of my own, which, as you can see from the stream of consciousness style of the Wiki page, may not be coherently expressed at the moment, but at least I’ve started to write them down.
The basic premise is “The web is an open world model, yet the majority of our applications assume they’re interacting with a closed world model (a single file or database), how do we rectify that, and what advantages might that bring?”.
The relationship between this, and the comments that Leung points to, is that both end up converging on application design. Choosing the correct language substrate is one design decision. After all you don’t want to be fighting the language all the way. But I think highlighting the closed/open world model distinctions can lead to application improvements whatever the language.