Baa Camp 2008
I just got back from a weekend at Baa Camp (photos at Flicker). It's been great time! If you are interested in more than my personal notes, Kim Hill showed up and did a three hour show with some of the attendees (the website says that audio archives of the show are only available for four weeks).
The Technological Singularity
- This was a long-lasting and wide-ranging discussion on singularities, mostly focusing around the idea of a technological singularity.
- My understanding of a technological singularity is that machines become able to create machines better than people, at which point we reach an impossible to predict rate of change.
- There was a lot of discussion about if we had lived through singularities previously (eg. the agricultural or industrial revolutions).
- The government of Estonia has decided to provide OpenID credentials to all of its residents. Sounds very interesting, but it does make me wonder about the security concerns around OpenID, especially around the known phishing issues.
- I'd never heard of i-names before, and I'm not sure I get the point, it seems to be just another implementation of a name services ... and I already have my own domain to solve that. I'm not sure what value getting an i-name would provide?
- One security concern which had never occurred to me is that if the domain you use for your OpenID URI expires ... anyone can be you. So if I ever let shand.net lapse, then whoever registers it next can be me to any site which I've logged into.
- I had a few interesting conversations with Marek Kuziel about the stuff he's doing, including writing his own Django based OpenID software and trying to OpenID enable New Zealand.
Zef and Lulu
Zef and Lulu gave a great talk about how to use visual tools to communication with groups. It was mostly tangential to my needs, but it was very interesting to see how they approach a group to get them involved in the design process. I particularly liked their "uber charts" which they use for plotting what a group thinks about something currently and what they think the eventual goal should be.
Open Data, Open Government
- When data sets are made publicly available, interesting things happen. For example, a company in Germany has apparently combined average weather, roof area and angle, and GIS data to figure out which sites it makes sense to put solar panels on. Things like this can't happen until the data is available.
- There are sovereignty issues with data, for example government departments using overseas web services for storing / managing data (eg. Basecamp) are being told to bring their data back to New Zealand, so it can be protected by New Zealand law.
Wiki & Publishing
- Adam Hyde has been working on a great project called "FLOSS Manuals".
- The manuals are largely being written by artists who are using the various pieces of software as part of their work.
- One advantage of this is that artists are often trainers or run workshops, so have excellent communication skills.
- Their website is based on TWIki, and they are in the process of releasing the various plugins they've created to make the process work.
- There are some other similar tools for using wikis as publishing tools. A plugin for PmWiki called Wiki Publisher and a Media Wiki project called Wiki Educator.
OpenID and Privacy
- People are interested in using OpenID as an authentication method for "high value" transactions (eg. logging into your bank). The question posed was what would have to happen to make that feasible.
- There was a lot of confusion about authentication vs. authorisation ...
- To me, it seems that in order to make high value transactions feasible via OpenID, you would have to vest some entity with the authority to vouch for people (eg. the government or Verisign). This almost completely removes what I love about OpenID, which is that it gives me the ability to control my own data.
- If you don't vest a third party with the authority to vouch for OpenID's then you're back to the web of trust which is never going to fly with the banks. Even if it did fly, it would necessitate a publicly query-able social graph, which has all sorts of scary privacy concerns (eg. insurance companies not giving you health insurance because one of your friends died of a heroin overdose).