Spamassassin was failing, all I needed was a kill, zap, and restart. But in my typical kid-in-a-candyshop fashion, I used it as an excuse to bump up the server to the latest goodies. BAD admin! But it wasn’t very many packages, and gentoo made it smooth as always. Included a bump to kernel 2.6.25-gentoo-r8, and a postfix bump (which dispatch-conf made trivial). Sweet.

The simple example will show you all you need to know. Handles all those crazy command-line options for ya, nice.

Official download page
Homepage

0.5.11 was reviewed and approved, which makes it public. This is great for two reasons:

  • Users can install the addon without logging in to mozilla.org
  • Users will get live updates to the latest approved version

This is VERY good stuff. To put it in perspective, I had a couple hundred downloads this morning, and I’ve been watching that number jump by 100 every hour or two – it’s currently over 600. Exciting times for my little addon!

I released 0.5.12 by request to address more international google URLS (eg, google.fr). The nice guy that requested it even donated $10 afterwards. A good day all around. :>

Avelsieve is a really nicely done web interface for email rules. It is a plugin for Squirrelmail, a tried-and-true webmail client, that works with a cyrus IMAP server, another old dog that is a nightmare to install but gets the job done no matter what it is.

The system has been handling my email for years, but for the past few months, I haven’t been able to get Avelsieve to edit my email rules. It’s very picky about compatibility. Today I finally dug in and fixed that. Apparently cyrus updates broke avelsieve, but there’s a workaround. Turned out to be one line in the avelsieve config file, updated as follows:

# MDM Will this fix the problem of not decoding scripts?
# MDM YES!  IT DOES!  Thank goodness.
# see http://thread.gmane.org/gmane.mail.squirrelmail.plugins/14106
#$avelsieve_disabletls = false;
$avelsieve_disabletls = true;

Yay! My pretty and powerful email rules are back!

Official download page
Homepage

Google updated their results page today, Google Results Walker has been updated to match.

It was a simple change of the footer, from using an ID of ‘footer’ to using a className of ‘clr’ (?), but it was enough to break walker. I updated the code do more checking of assumptions, so next time the placement may get thrown off but it won’t break.

Have fun with it!