Tim O'Reilly and the three 'C's
Well, I made it to FOSDEM this morning with time to spare. Bumped into Sarah and Josette setting up the O’Reilly book stand. Dave Cross is here, as are Leon Brocard and Richard Kilmer, whom I snapped during Tim O’Reilly’s keynote “The Open Source Paradigm Shift”.
Tim’s talk was very interesting, especially coming at this stage in Open Source’s lifetime (early on). He mentioned afterwards that he’s given that talk a few times now, and people are starting to catch on to what he’s saying. What is he saying? Well, it’s nothing particularly radical, nor is it anything that’s not been discussed before by Tim or others. But what was great about the talk is the way it put all the pieces together, and provided the audience with a view above the parapet.
Tim talked about how the focus is, or should be, moving away from software as a product, and further towards being a commodity. Pieces of software become merely components in undertakings that are larger than the code itself. Citing the usual suspects (Amazon, Google, EBay), he pointed out that what was important today was:
- **C**ommodity
- Software is a commodity. Licencing becomes less important where open source software doesn’t get distributed further, whereas the applications that build on that software are the things that get ‘distributed’ (across the Web).
- **C**ustomisation
- People inside Google are the people inside the black box, inside the [Mechanical Turk](http://www.theturkbook.com/preface.php&e=7317). They’re the people who make the difference between a piece of software, and a software-based *service*. They make the applications *dynamic*. Remove those people from Google, and leave the source code. Do you still have Google? No.
- **C**ollaboration
- Usenet was arguably the birthplace of collaboration. What it’s become today encompasses many things. Adhocracy, social software, users contributing to Amazon with their book reviews. Napster brought automatic participation through its architecture, making it easy (compulsory!) for users to participate. “The architecture of participation” is a nice way of putting it. So with a focus on collaboration and users contributing with information, we have to be careful of where we put data and what the terms and conditions are. (I immediately think of [Orkut](http://www.orkut.com/), and the nasty business of what happened to [CDDB](http://www.cddb.org/).)
“Interoperability and open data formats may be more important than source code availability”
It’s something I’ve said many times before: what’s perhaps even more important than open source is open protocols.
(More pictures from Fosdem 2004 here.)