I have to warn you - I've been (together with Pavel Hak) an official organizer of this year, so most of the following stuff should be taken as potentially biased and subjective - not that I want to paint it pink or so, but the organizer's point of view is naturally different. And when it comes to the following year, the ideas presented are mostly just wild unofficial speculations as there's not even an official team of organizers yet.

All comments regarding this year and ideas for the following year are welcome. Send them directly to me or even better to our official list at Google groups.

How was this year ...

First a bit of statistics. There were about 110 participants, 95 of those were regular attendees and the remaining 15 people were speakers and organizers. From those 95 attendees, about 30% were students. Regarding those numbers, it was almost exactly the same like last year, and unless we significantly change the conference we can expect about the same numbers next year.

There were 8 regular sessions, which is about the same like last year (not counting the short presentations of sponsors and short keynote for the last year). So no big difference here, too. But that's kind of expected for a one-day single-track conference. The number of sessions for a single day simply is 8, maybe 9 if everything goes perfectly OK (which is only rarely the case).

We had two guest speakers (Magnus Hagander and Marc Balmer) this year, each with one session. That sounds like a reasonable compromise for a conference organized for local developers in the first place. Just a quick hindsight - in 2008 we had one guest (Peter Eisentraut) with two sessions. In 2009 we had three guests (Magnus Hagander, Simon Riggs and Kristo Kaiv) with four sessions. And in 2010 we had one guest (Kristo Kaiv) with one session.

So I guess two guest (i.e. two session) are a reasonable compromise that does not break the whole "local conference" concept and yet it allows to bring some really interesting guest.

The truth is that when a guest was presenting two sessions in the previous years, it was usually a quick stand-in for a speaker who couldn't come for some unexpected reason. That fortunately did not happen this year, everything went perfectly (expect for the delays in schedule, but that was somehow expected).

An unpleasant shift compared to the previous year was the decrease of sponsorship subsidies. Next year we have to act much actively and sooner, so that we know how much we gonna get at least a month or two before the conference as that information allows us to set the entrance fees etc.

But we were somehow lucky, as the conference was held at Charles University and they provided all the necessary stuff (room, beamer, ...) for free. And we have not used any agency for the first time, the whole event was organized by the newly formed User Group - that saved us a bit of money too.

Another shift - this time a pleasant one - was the level of catering. Instead of sandwiches bought at a supermarket, and very limited level of catering (as in the previous years), we have decided to pay for a regular catering. It brings higher expenses, but if we want to constantly improve the overall impression of the conference, there really is no other way.

... and how might be the next year

When preparing the conference and during the conference itself, a lot of very interesting ideas emerged. We haven't used them for this year but I guess we'll reconsider them for the next one. Let's see some of the more interesting ones.

Workshops

The first idea is to organize workshops, i.e. about 4 hour session where the attendees get acquainted with some new or less knowh functionality, under the lead of a lector. The PgDay 2010 was a big inspiration in this case, and I guess the workshops there were a big success.

Length of the conference

The second idea is related to extending the length of the conference, by either adding parallel tracks within a single day, or using two days with a single track os sessions. Sure, the necessary condition would be to collect enough interesting session, so if you notice something interesting (interesting system built on top of PostgreSQL, an interesting feature etc.) just let us know.

Adding workshops would definitely mean another day - a regular conference on the first day, workshops on the second day.

Recodring the sessions

The last idea was to record the sessions using a camcorder, which was repeatedly discussed during the preparations for this year. Several people who could not make it to the conference because of other duties asked us for this. We're going to publish slides, but we all know that the slides without the actual comments from the speaker are often hard to follow.

From a technical standpoint, recording a session is not a problem - it's nto a hollywood block buster so all you need is a reasonable digital camcorder, a way to record the signal from a microphone and a bit of work to cut the recorded stuff. I do this once a month for a local JUG (see the youtube stream) and it just works.

The problem is not all speakers want to be recorded - some of them want the use the session for other conferences (and the video somehow complicates that). And some of the speakers are not really relaxed if the session is recorded.

Another bad thing is it might impact the amount of money received from entrance fees - we'll please several people who could not make it to the conference, but others won't come because they can watch it on the video. Which means you'll get less money from the entrance fees before the conference, and that's a rather portion of the conference budget (this year, we've paid about 30% of the invoices from entrance fees).

We have not ruled out the recording for next year - but if we decide to try it, each speaker will have the option not to be recorded, and the recordings won't be available immediately and it probably won't be a free service.