A blog entry by Charles Curley describes how to store the nanoblog contents in git, and lists what files can be left out of the repository since they are generated by the nanoblogger script.
I'm writing the blog on my laptop and I'm serving it from the web server. In the interest of security I would like to 'publish' as few files as possible, to minimize the attack surface. As such, I'm using this publishing command line:
BLOG_PUBLISH_CMD='rsync -e ssh -arz --delete --exclude-from dont_publish $SOURCE/bit_on_the_bus/ $SERVER:$DESTINATION/bit_on_the_bus/'
The contents of 'dont_publish':