Here are my notes from Saturday's event.
NOTE:this is a rather long and link heavy post, if you were present and notice any errors or omissions let me know (via the comments)
The day started off quite slow, but once we numbered something around 6, the event kicked off.
As we were such a small group, a traditional unconference grid seemed overkill, so we went with a more, "does anyone have anything on this topic?" approach.
First up was
Teemu Harju who gave a brief overview of
PubSubHubbub. In Teemu's opinion XMPP was not that much harder to code for than HTTP, it's just different.
Teemu saw PubSubHubbub as very important contribution for the move towards the real-time web. He is also working on his
own PubSubHubbub server written in Erlang. Teemu also pointed out that the PubSubHubbub definition was only at version 0.2 currently, and still a project dominated by Google Engineers.
As an aside, due to most of the attendees being non-native English speakers they had not realised that
hubbub was a real word.
As PubSubHubbub was a Google technology, this brought up the question of "What is the use of Wave?".
After some back and forth conversations, it seems that still for many people Wave is attempting to solve a problem they don't think they currently have.
One good thing to come out of the conversation to use the example of
Etherpad to demonstrate the basic use of Wave without having to actually use wave. A useful tip as invites are still relatively limited/restricted
The next topic was
OpenID.
The main point from this session was around the usability of OpenID. Once again there seemed to be some confusion around the exact problem that OpenID is attempting to solve.
The use of a URI to login rather than the more traditional regular username or email style username, seemed to be the main sticking point.
One important point was that OpenID would be preferable for Government organisations as it is not tied to a single commercial entity (like Twitter or Facebook)
Although, now the large user communities (Google, Yahoo AOL etc) are all OpenID providers, there is still understanding gap for more mainstream users. The use of a Facebook login was considered more understandable, although most of the people attending had never attempted to use Facebook connect. The actual use of FacebookConnect was at least as poor a user experience as OpenID for me.
One of the attendees (I think it was Aapo Rista) was involved in a project that was building a site based on
Django and also using
Trac, but was having problems with OpenID. The good news was that during the day he was able to get a working demo up and running, and at least from the demo he gave it seems the Django plugin makes use of OpenID easy.
OAuth was next up. Everyone had heard of it, but all the examples initially were just around Twitter.
Luckily as
Kai Lemmetty of
Floobs and
Arto Luikkonen (who knows the API well) showed an example of using OAuth using the Floobs API.
The next major topic was
Open Micro Blogging. Although there was not much interest in this topic by itself, it sparked an interesting conversation about how in places like Africa, services are being built on top of SMS. The concrete example given was of the Masai learning of cattle prices. Also, that this information was increasing the desire of the local peoples to learn to read and write.
As
Petri Kola was present (and had just finished two days of rather intense discussions around the topic of Openess at
OPEN Symposium) he gave a short presentation of the work he had been doing around open data and getting open access to Government data.
There was also a short session around our generous hosts for the day
Hub Helsinki.
As a final note i have to say thanks to
Jukka Pellinen for all the help on the day with extra equipment.
