There was more unintentional downtime this evening, sorry about that.
Over the past few weeks I thought we had nailed down what was causing these periodic crashes, but obviously not. Will keep on searching.
We were offline this evening from around 7pm to 9pm.
During a crash we generally update the Facebook page so that you know the status:
http://www.facebook.com/efswitzerland
For those of a technical interest, we're running on an dual quad-core server (i.e. 8 CPUs) and usually during the day we're at 50% to 60% load. At night that drops to 15% to 20%. When it spikes to 100% or more then things tend to fall over, the database gets corrupt, and repairing it takes about half an hour.