I attended SWIB15 in Hamburg, this is a very brief write-up (I’m no longer contractually obliged).
This year’s SWIB made a welcome return to Hamburg, which is a lovely place.
I spent Monday at an excellent workshop with Tom Johnson on RDF.rb & ActiveTriples (the former I have used quite a bit, the latter, I haven’t). In the former case, Tom showed us a lot of things that are quite difficult to grasp and clarified a lot of things that I struggled with as well as generally introducing a lot of new information. We managed to hold tact with copy-coding the well-chosen examples too, which I think helps understanding a lot. ActiveTriples were similarly covered in the right amount of detail and I’d feel confident to use this technology in a new project that required Ruby on the background of the information given here. I am not a Ruby programmer although I use it quite a bit — and my colleague has never used Ruby — yet we both came away with what we feel was enough to know what & how we can do things in RDF.rb/ActiveTriples. Tom deserves credit for continuing the strong tradition of SWIB workshops — especially as they’re bloody marathons for the workshop leaders.
The only downer on Monday was getting stopped by aggressive, unco-operative HVV ticket inspectors for having the wrong ticket — an honest mistake, but one that resulted in a ruined afternoon. It wasn’t the fine — a pittance — it was the attitude.
Tuesday was a day ruined by my own talk — presenting isn’t an activity I like, I like a discussion where I’m not monologuing. Stress & nerves made everything else a blur. Although this was the case, I appreciated both Simeon Warner and Fabian Steeg’s presentations in my session as they both provided interesting information and a welcome distraction from my own impending doom.
Wednesday had a couple of presentations that I would have attended the conference for alone — Karen Coyle giving a tour-de-force on FRBR and Tom Johnson (again!) giving perspective on records in the linked data reality.
My general impression is that we’ve now reached a stage where we’re doing actual read/write linked data or plain conversion. Those of us doing read/write are in quite a different place than those doing conversion. These are exciting times and I feel that the next few years will see linked data in libraries either made or broken.
Technologywise, there is a lot of convergence on JSON as a format of preference; I’m not sure that this is a good idea, I have issues with storing data as strings — as I said at the conference, I believe in expressive data and less intelligent software. We use JSON extensively in Oslo, I’m not sure we share the same joy that others feel.
There are some exciting projects going on out there, which I hope bring fresh perspectives to bear on cataloguing and presentation as they mature. I suspect that the team in Oslo stands at an extremity in terms of how we view linked data and the semantic web generally. I’m not sure that this should be important — as we learnt in Trondheim and Oslo, there is a heavier price to pay for methodology choices than there is for technology choices.
As usual, talking with everyone — hat-tip Fabian, both Adrians, Felix, MJ, Tom, Niklas — provided new information; I have started applying some of this in my work this very day.
Thank you all for a great conference — especially the poor canucks that had to put up with me at dinner!