I help with Melbourne Perl Mongers.
I spend an awful lot of time talking about Perl, and have had my picture in the Australian newspapers with a camel. That's rather scary.
Having swiftly recovered from the dreaded Linux virus and having got most of my work back under control, I figure it's about time that I finished my thoughts on linux.conf.au (LCA) 2006.
I've never been to an LCA before, and as such I'm unfamiliar with the idea of a mini-conf. The general idea is that time and facilities are set aside for a small, independent conferences to be held on particular topics. And I mean independent; the mini-confs have their own organisers, paper committees, webpages, timetables, and discussion topics.
The mini-conf idea is a good one to allow more content to be packed into the conference time, but without much extra work for the main organisers. Less work for the main organisers is a Very Good Thing.
The downside is that as the mini-confs are independent, they're not necessarily co-ordinated in any way. This is fine if you pick a mini-conf and stick to it, but it's more problematic if you wish to move between mini-confs. Talks may be scheduled into different timeslots, or run over- or under-time, making clean transitions difficult.
There was more than one occasion where I had to break conference etiquette by leaving a talk that was running over-time to try and reach another I was interested in. One of the more frustrating experiences was trying to figure out timetables; one has to check half a dozen different timetables on different sites to know what's being presented.
Despite my trying time at mini-confs, I think that organising a Perl mini-conf for LCA 2007 would indeed be a great idea. This year there were a few mumblings but a Perl mini-conf per se didn't happen (although we did have a Damian mini-conf).
I'm pleased to say that the conference got better as time went on. To my relief the main conference was much better co-ordinated than the mini-confs. However there are some faults that I kept spotting, although these are common to many conferences, not just LCA. My advice for anyone involved in a conference would be this:
Speakers: Control your audience. Don't let them interrupt you. Don't allow them to start long discussions on tangential topics. Everyone else in the audience is present to hear your talk, not the opinions of the person in the second row. Ask for questions to be left until the end.
Room monitors: Control your speaker. They will ramble on the end for as long as you allow them. Talks that go overtime are unfair to the next speaker, and are unfair to any audience members who wish to change streams.
The other thing which amazed me is that LCA doesn't have conference proceedings! I kept finding myself wanting to check the proceedings for references and details of a talk that I had just attended, or examine the proceedings to plan which talks I'd like to attend the next day. Proceedings are a wonderful thing; not only do they satisfy delegates' cravings, but they also help improve the quality of the speakers by making sure materials are prepared well in advance.
NZ Perl Mongers
The best part of the entire conference was being able to catch up with a great many New Zealand Perl Mongers. Apparently Wellington.PM is quite alive and active, and it seems that New Zealand is using Perl for all sorts of interesting things, including running their DNS and electoral roll.
The conference dinner was incredible. The food and venue weren't anything to write home about, but the auction was breathtaking. $10,000 AUD was raised towards the establishment of the John Lions Chair in Computer Science. This was matched by a further $10,000 from Linux Australia, and the whole bundle with be matched by another $20,000 from USENIX. $40,000 AUD is an impressive sum to raise at a single dinner.