Short, social, OSCON notes.

May presenters were influenced by the Dick Hardt presentation from last year.  At least the keynote speakers had all studied the art and attempted to break the monotony of bullet points.  To me though too close a mimicking of Dick's style (inspired in turn by Lawrence Lessig) is poor. Some got it right, some were embarrassing.

Question Copyright

I really liked Karl Fogel's 45min session, despite him running out of time. The presentation hangs off the page above. He's studied much of the history of Copyright  (the Brits cursed America with it before independence).  The above is a must read for all concerned with media, open source etc etc in the modern age, principally debunking the various media industry efforts to artificially link "credit for a work" with "copyright".

Five a day

Robert 'r0ml' Lefkowitz gave a very funny keynote (many meta-meatphor jokes) linking Open Source to attempts to get us to eat five pieces of fruit a day.  No deck as yet:

History of what we are doing now. Helping future archeologists.

Jason Scott - really interesting keynote with a deeper message for technologists who just might be doing something. Apart from the documentary on BBS pioneers and FidoNet (the pretext for his keynote) that is available under a liberal copyright, I hope his final request is more widely broadcast:   keep notes, take photos and archive backups for the long term that well help people doing historical analysis of our recent past.  There's also I guess some benefit to patent busting as prior art.  See in case something is uploaded.

Google's 'service' announcement.

Brian Fitzpatrick has been too busy at the conference itself and in his hotel room coding to do much drinking. He suggests that its been work on the enigmatic thing that Greg Stein will announce later.  I have hopes on what this may be seeing as it involved a whole bunch of Svn / DAV committers thinkers. We'll see.

2pm update.

So, it's a nascent SourgeForge knockoff. Simpler of course.

It has Subversion with a better backend based on 'Bigtable' (I wonder if that code will be donated back to the Svn codebase - hosted at Tigris which is affiliated to Collabnet).  There is also a smart label based issue tracker. Its appears to be inspired by a simplify Bugzilla, avoid Scarab and be inspired by Trac. Isn't referent to Jira at all. The issue tracker was over-demoed in the session.  It leverages Google-groups for the mail-listy side of things which is the second reuse of existing infrastructure (mandatory Gmail was the first).

There are only seven licenses to choose from during setup. They don't want to encourage license proliferation.

There is nothing here so far that, Collabnet (their hosted product or Tigris or can't do.  It  is what comes next for 'Google Code' that's key.  My 'hope' for Stein's baby isn't met yet. I wonder if it was just cut from the initial release.


July 27th, 2006