Octopress was built on top of Jekyll, so I figured the move would be fairly easy. On the whole, it was. I was able to move the content over without too much difficulty. I ended up having to back out some of the Octopress plugins and either write the functionality myself or look for alternatives. Thankfully none were a terribly big deal.

The only major change was to the overall structure of the site. I opted to move my blog and linkblog into Jekyll collections, which worked out pretty well. I opted to go with the newer jekyll-paginate-v2 gem to build out the archive pages for each and used jekyll-tagging to generate the tag archive pages. In the process, I also moved all of my post “categories” over to being “tags”, since that was more appropriate to how I was using them anyway.

It took a few days to get it to the point where I feel I can merge the Jekyll-only branch back and publish for the first time, but I think it is pretty stable. That said, there are likely some bugs yet to be discovered. Please let me know if you find one; more sets of eyes are always better.

In case you’re curious why I’m making the shift, it’s pretty simple:

• Octopress development seems relatively stalled;
• Octopress includes dependencies I don’t really need;
• I’ve been writing a lot of Jekyll plugins and maintaining them without using them on the latest Jekyll is challenging;
• I’m comfortable writing my own Rake commands; and
• This is the big one: Incremental builds.

I’m hopeful this shift will make it easier for me to get content published more quickly and more reliably. I may even move my build processes into Travis CI to further offload the work. But I’m getting ahead of myself… let’s see how Tuesday goes first.

Jeremy posted in detail about how to set it all up and provided the PHP code he’s using to make it all work. As I’m running a static site on Octopress, I ported it to Ruby as a Jekyll Generator. I’ve posted it to Github, so you can grab it there if you so desire.