» 22 Sep 2008
After setting up my blog with hobix installed on my Dreamhost account yesterday, I realized that I wasn't dreadfully fond of that situation either. Ideally I should be able to write content in the most convenient manner possible and have it somehow get to the interwebs. I'm much happier using my local install of Vi and having full control of all the libs that bake my blog into a delicious website ready for dissemination. I am also happy when my data can survive me randomly overwriting critical things.
Given these general desires, I set out to modify hobix to run on my laptop and then, using git and github, get the compiled website to my webserver. Hobix seemed to operate in a very simple manner than would make it easy to modify, so I forked it on github and got to work. I quickly added in a hook and configuration option to commit and push to the remote repo specified by @blahg@. I then hooked that up to a repo on github. When changes are pushed to github, a post-receive message is sent to a PHP file on my webserver that then performs a @git pull@. All the plumbing is working now.
While I was hacking this together, I ran into some issues that required going deeper into the code of Hobix. I ended up completely deleting the lockfile capabilities and also changing some of the path handling code that screwed up the templates when generated on Windows. While I was going through this, I realized just how much code was there. It seems to be an extremely fully featured and complicated piece of blogging software. I'm not entirely sure how I feel about that. I'm getting somewhat burnt out on blog features. I kind of feel that I should be able to compile my blog using @make@ like god intended. If I get some more free time, I will probably end up trying my hand at a Rake based blog build system.
I really like my current setup and I think Hobix is quite the tool. I will continue using Hobix while I think of ways that might make me happier about my blog system.