Chris Lord has posted an essay entitled “Introducing Social Calendaring” which I think is an interesting read for Bongans. It took me a couple of reads of the paper to actually figure out what was being proposed, but that’s probably because I skimmed it first a few times with my own preconceived idea of what was being proposed.

I would love to see some of these ideas implemented in Bongo, although I don’t know to what extent some of it would work. While I like some of the proposals in the paper, it is a shame that there was no kind of implementation work, because there are a couple of things I really think deserve more analysis (this might not make too much sense until you read Chris’ paper):

  1. the proposal is mainly decentralised, offering a peer-to-peer system for sharing of calendaring information, but without much discussion of availability of data access. Peer to peer works really well where people are sharing the same data, like music, but I’m not convinced about calendar sharing. Unless you end up sharing other people’s data for them (and vice versa), I think you end up relying on people being online at the same time in order for sharing to work well.
  2. synchronisation for “non-primary clients” seems antithetical to the idea of decentralising the informaiton. In particular, if the availability problem was solved, you wouldn’t need a separate method of synchronisation, in my opinion.
  3. the tagging is extremely important in my opinion, and it would have been nice to see more discussion paid to the idea of RDF usage. In particular, I didn’t find the UI analysis terribly thought-provoking, and Chris mentions that developing a good UI on top of an RDF-based system would be challenging. Some further description of that challenge would have been great. I’m not sure an RDF system would be the way forward compared to tagging, though.

I’m hoping to read the paper in detail in a couple of weeks, and while it’s not something which is relevant for Bongo 1.0 (sadly) I’ll hopefully distil some of the salient points into the wiki, and perhaps look at some of these questions myself. I’m particularly thinking about the “thick client” type solution Chris talks about – where data storage is primarily local for offline access, etc. I’m not totally sold on that – for example, I think SMTP has a good balance of decentralised peer-to-peer design without suffering the availability problem, and that’s one of the (few) things I like about iTIP.