BarCamb 3

This weekend has mainly been filled with BarCamb 3. I had a modest part to play in it’s running (watching doors, carrying stuff) but still had plenty of time to attend a number of interesting sessions.

This will be my third BarCamp event and it’s nice to attend one on my home turf. What I like so much about the format is the shear serendipity of discovery. While a lot of the speakers have some sort of idea about what they will present there is plenty of scope for last minute presentations or contributing to sessions like Lightning Talks.

So far I’ve attended a number of interesting talks/discussions. The first one was a discussion about the relative merits of micro vs monolithic kernels. This led onto discussion of if the benefits of isolation of components could be achieved with a JITing Virtual Machine and if that would solve some of the performance problems that micro kernels traditionally suffer from.

The next talk was an interesting history of Babbage’s Difference Engine including a number of the implementation details. The talk included a demo of the engine written in Flash. Unfortunately time didn’t allow for delving deeper into the “Method of Difference” from which the engine derives it’s name.

Another interesting talk was Eva Amsen‘s discussion about scientific communities and the un-conference concept. It started with discussion about the plurality of science social networks that exist and the problems that fragmentation causes. This is slightly surprising given scientists were one of the first groups to use the ‘net extensively for collaboration. The discussion then moved on to musing about why scientists haven’t taken on to the BarCamp style meet-ups. A lot of practical issues surrounding not announcing research before publication and worries about presenting partial data and work in progress. Events have been shown to be successful (by for example inviting them to a “normal” conference and then cancelling the schedule once they had arrived). There also seems to be a thirst for normal tech people to meet scientists which may not be reciprocated by the science community. I left the discussion feeling a little sad that the ad-hoc collaboration that is so natural in the Open Source world still has some way to go in the realms of science.

All in all it’s been very stimulating so far and I’m looking forward to the closing sessions this afternoon. There is going to be some 80’s computer nostalgia which I’m quite looking forward to.

It seems to be the season for BarCamps as I’ve got another one next weekend when I’m attending OggCamp 10 up in Liverpool. Anyone want to join me?

One Comment

  1. Definitely wished I’ve have dwelt a little long on the method of finite differences. My slides showed an example of how the differences were used to determine the starting values of the difference engine.

    I just felt I could have explained that a little bit more clearly that determining the differences meant that the next value in the pattern could be calculated by using only addition.

    Consider x^2

    x: 0, 1, 2, 3, 4
    x^2: 0, 1, 4, 9, 16
    diff1: 1, 3, 5, 7
    diff2: 2, 2, 2

    There are 2 methods to calculate the next value in the pattern:
    5 * 5 = 25 or
    2 + 7 + 16 = 25 (This works because the previous result 16 is stored in the machine along with the differences 7 and 2)

    My slides are going up as soon as slideshare starts behaving and I plan to upload the simulator at some point.

    Thanks for the feedback.

Comments are closed.