Software packages, but it's not a tree, it's a graph!

By | August 30, 2012
This is exciting. We're moving from objectized, linearized data into complex data ecosystems, where data is represented by webs as opposed to trees. The same is true of software, and most likely everything is going to jump from a linear, categorized form, to a graphy ecosystem form in which relations between arbitrary objects may be created, and broken, on a daily basis.

And that's what is so significant about Debian's move to RDF: RDF is leading graph definition format at the moment. And Debian is one of the biggest Linux bases. A big sign that the the paradigm shift is coming: Welcome to the Interwebz.

/via +Michael Hausenblas 

Reshared post from +Edd Dumbill

Debian package metadata now available

This is really exciting. The AMDS.SW vocabulary is an EU project, building on some of my earlier work with DOAP (Description of a Project).

You can now get machine readable package descriptions by using content negotiation on the Debian package tracker, e.g.

$ curl -L -v -H 'Accept: application/rdf+xml'

(or just add '.rdf' to the end of the URL, e.g.

This means one of the largest collections of free software is now computable. Great work by Olivier Berger.

Debian Package Tracking System now produces RDF description of source packages | WebLog Pro Olivier Berger
Here’s a second post on the subject of RDF descriptions for Debian source packages (see the previous post for some context). From now on, the Debian Package Tracking System (PTS) will produce, alongsi…

4 thoughts on “Software packages, but it's not a tree, it's a graph!

  1. Abe Pectol

    I should note that `apt-cache dotty` been there for quite a while already. And, also, that most graphing software is barely able to draw it at all, never mind positioning the nodes in it neatly. (gephi somewhat succeeds at that, though)

  2. Sophie Wrobel

    +Philipp Sengelmann I'm quite sure that this is not about database optimization – especially given that Debian source package description database is nothing compared to the scale of large databases. Rather, it's more about information accessibility to things outside of the current ecosystem. Typically, package tracking is done by the package manager, and no one else really cares about it. But moving it to RDF format means you can have new ecosystem members – for example, a search engine or social network – start to interact with the data contained in the RDF descriptions. Not that it would make sense for a social network in classical sense of a social network, but that's exactly the point: some completely unrelated system can easily interact with RDF data with relative ease. And that's what makes data prepared for a big interconnected ecosystem.


Leave a Reply

Your email address will not be published.