Using Git to Maintain Your Website

Hmmm...interesting, but it seems like this takes automation a bit far. I also use Git as a single developer on a project of mine, and I have staging and production environments on my server, but I have simple deployment scripts that I run to update the changes on the server, clear any cache files, etc. It's not that much more complicated than what he does here, but it gives me the peace of mind of knowing that I can commit my changes without the world seeing them automatically.

I used to use svn, but these days I use git for maintaining static websites.

Maybe I missed something in this article, but the key to static websites (as far as I see it) is using a templating system. Put your templates, content and configuration under version control and you're pretty much done. The website itself is a generated artifact.

Or, use ikiwiki (http://ikiwiki.info/). It takes your git repository, a directory hierarchy of Markdown files, and will automatically compile them to HTML each time you push. Even more elegant.

I do something similar with my site, but without the extra intermediary repository. On a push, the server-side hook simply hard resets the content of the folder to the tip of a specific branch. The website itself acts as the canonical repository of, well, itself.

http://yangman.ca/blog/2009/08/06/website-management-with-gi...

these days I use github+github pages, even although I wrote http://drydrop.binaryage.com ;)

The Ruby Best Practices blog (blog.rubybestpractices.com ) uses a a git-based system that's quite nice.

http://github.com/sandal/korma

I wrote up some background on it here: http://blog.rubybestpractices.com/posts/jamesbritt/2009-04-1...

I have a very similar setup with post-receive deploys on remotes "stage" and "production".

git push stage master

click around staging environment to make sure nothing's broken

git push production master

you're live.

Using CVS to Update a Web Site

http://www.kegel.com/cvs.html