simonb.com: a 35mm comedic short about a man

Note to self: on Gentoo postgresql configuration files (pg_hba.conf and postgresql.conf found in /var/lib/postgresql/data) need to be owned by the postgres user. So if you edit them as root, reset the owner:group before restarting postgresql.

So I’ve just done my regular emerge sync etc. It seems that everytime something new is installed printing stops working. The current issue was fixed by adding a line in cups.conf which explicitly tells it which network address:port to listen on. Why this worked before we’ll never know.  So here is the saving line:

Listen 192.168.1.2:631

This weblog is (currently) running ExpressionEngine. I just had a bizarre problem: when I went to the home page it was only displaying a message saying “Invalid URI” Sometimes it would load part of the page though without the css. An intermittent problem - the hardest to track down. I searched around the forums and the net for why this could be - found nothing. Nothing changed as far as I know so it was a bit odd. There was also nothing obviously wrong in the logs. I fixed it by restarting Apache and MySQL - maybe only one of those needed to be restarted in retrospect. I suspect a bit of PHP had gone wrong somewhere.

Update: Well it turns out that a simple restart was not the solution to the problem. EE’s “Invalid URI” response is not very helpful so I modified the php code to printout the actual URI. This is where it gets weird. Yes I recognised part of the URI but it was from a different virtual server which is using mod_python not mod_php and using Postgresql not MySQL. There seems to be a wondering pointer causing mod_php|python to overwrite eachother. I installed the newest Gentoo versions of Apache and restarted postgresql - it seems OK now…