I was thinking that I should do a blogaversary post. Today I looked to find out when the (in)auspicious date was, and discovered that it was January 9th. Forgot to set a reminder. This year, January 9th was the 770th post. So thats just under 2 posts a day, every day of the year. Wow. I didn't know I had it in me to write that much (500+ page books notwithstanding).

One thing that I've felt guilty about for most of the first year is that I haven't generated all that much original content. Julie has been much better about that than I have. Many of my posts are commentaries on other things going on the the computing world, which is certainly a part of what blogging is about. But I'd like to do some more original content this year. We'll wait and see.

Something that I've been the most happy about is the chance to "meet" new people and do some projects together. So I hope that I get the chance to do even more of that this year.

When you are new to Python, and I consider myself new after a year of working on pyblosxom and a few months on Chandler, you still end up flipping through documentation.

Today I discovered the command line pydoc tool, which give you man like functionality for Python. This is pretty useful, but not as useful as it could be. The best man tool that I ever used was tkman, which was a Tk based man page browser, which also used glimpse to index the pages. That made it a lot more useful. It's to bad that glimpse has gone away, I constantly find situations where I could use it. Of course, I could use Lucene.... The snappiness of Mail.app's indexed search is one factor that convinced me to bag Thunderbird. The handling of threads (while not perfect) was another.

My other discovery was two functions related to info files in Emacs: info-complete-symbol and info-lookup-symbol, which work off of Info files. So I grabbed the info versions of the Python documentation, appended the directory to Info-additional-directory-list, and now I can complete symbols from the Python standard libraries, and look up their documentation from inside Emacs.

Today I started to do some work on the big RSS feed test that I've been working on for Chandler. While Googling around, I discovered this announcement of a method for using kcachegrind to give you a nice visualization of the profiler data produced by Python's hotshot profiler. It requires KDE libraries, so I decided to try to get this working on Linux first, being afraid to try and fink the necessary libraries for KDE. Turns out that there is a kcachegrind package for debian (in unstable). Unfortunately, that version doesn't include the scripts needed to make this work. So I grabbed the unstable version of kcachegrind, unwrapped it and used the script in there. Every thing seems to work just fine, and its a nice way to manipulate the profile data. Thanks to X11.app, I can at least view the visualization on the Mac. I'm not sure if I'm up for KDE libs on the Mac just yet.

On tip if you decide to do this. The input file to kcachegrind must be of the form cachegrind.out.NN where NN is a number. That was really a nasty pain to figure out.

Thanks to the hard work and generosity of Thom May, and the Gnome/Debian Planet Hackers, Planet Apache is now up and running. RSS feeds are available.