Of course, use.perl.org is a blog, and I have several (largely dormant) MT blogs. Let me be more specific.
I used to have a todo list in the form of a large text file. Then I realized it was getting out of hand and turned it into a Vim wiki (my google-fu fails to find the exact recipt I used, but viki is more or less the same concept. When I changed job, I forgot to transfer the config files across... so I tried the Vim outliner instead It's very nice indeed, and it lends itself to a date-based view (outline by year, then month, then day).
dialog at work was raving about blosxom (http://www.blosxom.com/) and it's really rather nice. With the wikiwordish plugin you can also start linking together your entries, categorizing them etc. and it all just works. I love the way that it just builds the files up from your filesystem, so that simple things like editing and version control are easy. And though the default template is rather noddy, I guess it could be quite easily hacked to use Template::Toolkit and it'd be even sweeter.
But as I'm using it as a todo list, I need to be able to search and edit old entries easily. And auto-commit entries when they change.
I think that what I want is a tree processing engine that can (exactly like blosxom does) walk a tree, collecting documents to be published to web. But the engine is generic, so I can also use it to generate tags files for Vim, or to let me search for "All documents categorized TODO" from the command line. And of course, it needs to be easy to set up replication from my local machine so that the webserver only gets files tagged as "public".