The decision came down to a few points: #### Cost To host my old web site I had been paying for a small Ubuntu VPS from RAMNode. Even though it was on the low end of their available VPS packages, costing $7 each month, I was ending up paying for more horsepower than this modest site ever needs. I could have downgraded to one of the tiny VPS machines, which cost $15 each year, but at a certain point maintaining a server and having even the minor concerns of keeping it up to date and secure just to host an nginx process seemed totally overkill. #### Git storage The promise of being able to use Grav’s flat file format and storing site content into a Git repo never quite panned out for me. Especially after the addition of Grav’s admin panel, storing only the /user directory in a git folder meant that there was a constant stream of out of sync plugins versions and dependencies, I felt like I wasn’t able to quite focus on the content on the site, rather than the plumbing. With Jekyll, I can keep the site sources and the generated site HTML pages without worrying about keeping things updated. #### AWS sandbox I’ve been trying to get into Amazon’s AWS for some time. This seemed like a natural way to edge into it by taking advantage of AWS’s free tier capabilities. I knew from a few other articles that it was possible to use S3 to host Jekyll’s generated static content for basically (and literally) pennies. With AWS as an IaaS, the only thing I have to worry about how much bandwidth my site uses; for my small bandwidth site, I expect to make my annual web hosting bills less than $10. Also, I will not have to worry about a VPS going down, any security issues, and so on.
Getting Jekyll up and running on a Windows machine is not as easy as it should be, as is often the case with these open source tools. I eventually just used the steps here to get it up and running. Now I’ve got a quick platform to just write posts, let Jekyll compile it to a static page each time I save my Markdown files, and then just drop the resulting files into my S3 bucket. Easy peasy.