I haven't got around to packaging OSSEC for Debian yet - mainly because I haven't decided how to handle the fact that OSSEC uses a server->agent model that depends on the generation/importing of unique keys for communication (not unlike Puppet with SSL certificates), from an automation/Puppet perspective.
Update: Aegir subfolder support is apparently now in Aegir 2.x proper. See the ticket
About 2 and a half years ago, a feature request was made for the Aegir project to support the creation/management of 'example.com/mysite1', 'example.com/mysite2' subdirectory Drupal sites.
The Debian package that I made for Kippo last year, was a few commits behind (though not by much!).
When recently firing up a fresh Kippo sensor, I realised there were some bugs in the postinst script on a fresh install - additionally, there were some other bugs when removing and re-installing Kippo.
These have been fixed, and the Kippo package is now in-line with revision r219 of the subversion repo maintained by desaster.
I had a situation whereby I had to protect a site with HTTP auth, but exclude a certain IP address or two from having to use HTTP auth (e.g loadtesting).
The problem was, the site was also behind a Varnish proxy. So I couldn't do 'Allow from (ip)' in the Apache settings, because the IP would always be the IP address of the Varnish server at this point.
And obviously, I couldn't 'Allow from (varnish ip)' because that's the same as not having any HTTP auth at all :)
Way back in November 2011, I did an unusual thing: I co-presented a talk about Drupal and Continuous Integration/Deployment at DrupalCamp Toulouse, except that I was Skyped in from my home town of Melbourne, Australia :)
The talk is actually mainly presented by Greg Harvey of Code Enigma, one of my clients, and I just provide - well, I don't know, comic relief with my Australian accent? :)
Last weekend (on the 15th January) I delivered a talk on Aegir at DrupalDownUnder 2012. I tried to do something we've not done before in Aegir demos: demonstrate basic usage along with more advanced, automated use of Aegir.
Overall I was happy with my talk - I ran out of time, as I was a bit too ambitious, but I really wanted to show the advanced stuff without alienating the newbies.
I've recently been working on a project for a client that required Solr 3.4 (which we implemented using these packages for Ubuntu 10.04)
Additionally, in production we've implemented two solr servers in replication (one master, one slave).
Having never done this before, I followed the example in the solrconfig.xml that was commented out, as well as this documentation.