Well this weekend my home partition filled up. Normally that wouldn't be a bad thing, except that bashpodder and spamprobe both couldn't write to their respective checkpoint files. I tried restoring the spamprobe sp_words file from tape, but alas, it didn't want to restore. So, I blew away that file, and started my spam-analysis from scratch. That's not too bad because the spam at this account has been marginal recently, and it usually takes about a week for the filter to start working properly once primed with spam. The part that was a little more annoying was bashpodder (a command-line podcatching client). I've loved bashpodder, and it's served me faithfully for many many months. Unfortunately, when it gets screwed up, it REALLY gets hosed, and starts to download everything in the RSS feeds. This proved to be a very bad thing, as it once again filled up the disk with gigabytes of mp3 files. Worse, there's no "catchup" function in bashpodder, so it has to download every file, without interruption, in order to not download the file again. While driving last night, I decided I would write my own podcatching client. (After all, how hard could it be, right? :) ) I had already downloaded perlpodder and determined that I needed about half of what it did (perlpodder checks the ipodder directory for new feeds that match a certain keyword. That's the last thing I need at this moment). So I stripped out the portion that parsed the XML feeds, and slapped it onto my program wget-queue.pl, and created a new podcatching client called piddlepodder. Piddlepodder reads the feeds first, determines what needs to be downloaded, and then queues the downloads to be downloaded by wget. It keeps track of what it downloaded as it downloads. It also includes a catchup option to sync the client with the feeds, so it can be made crrent without downloading megabytes of files. I'll be releasing the source soon after it proves itself, so if you're interested, hang on a bit.