Lately I had to realise that the OwnCloud support for SQLite is literally only for testing, even if you have a single user, despite what I have read out of the documentation (which states you could use SQLite for few user sites).

I had several issues with my calendars as well as my contacts (failed to sync, failed to import, etc.), which all vanished when I migrated/converted the database to MySQL. Luckily, OwnCloud has support for exactly that use case. Just invoke the following (given you have setup php correctly).

I just fixed an issue that kept the OS X Contacts.app from synchronising with OwnCloud when running behind a reverse proxy. For those who wonder, a reverse proxy is useful if you have a gateway http server that forwards http requests for web servers in a private network.

I just patched the installation of Collectd Graph Panel to make it work together with Collectd version 4.10 on FreeBSD 10.1. The value string in the Collectd Graph Panel did no longer match the (old) Collectd rrd data.

After editing Packet Filter rules you might want to verify what you just edited before enabling the rules. Here is what I usually do on my server to not screw up the whole installation and being able to quickly get back to my previous configuration.

The default boot delay under FreeBSD is 10 seconds. This can drive you crazy if your server is down.

If you happen to lock yourself out of your own remote system with no hardware access to the server you mind end up in the same panic as me tonight. Since it took me surprisingly long, I thought it might be helpful to write the steps down to get a zpool mounted on a rescue (FreeBSD) system.

Today I was working with a colleague and reasoned about some code over chat. We sent filename and line number frequently and I got the impression that this is another perfect candidate for automation.

Today I wanted to download the right kernel source code and found this helping blog post. However, I got the feeling that the url can be built automatically to avoid fetching the wrong URL. Use the following command instead and it should always fetch the appropriate sources according to your currently running machine: