End of an experiment

A little while ago I started using Jekyll to manage my blog. I liked the idea of writing in my text editor, using Markdown, and pushing posts to the live site with Git. I still like all those things, but I have switched back to WordPress because, for me, it is still the most convenient way of publishing a blog.

Why I went back

Jekyll is a sophisticated static site generator, but there are lots of common blog features that it doesn’t—and perhaps shouldn’t—support. A lot of these are simply natural features of a website that stores its content in a database.

In the end, it was the convenience of using WordPress that won me back. Jekyll is not difficult to use, but you need Git and GitHub or Ruby and FTP (i.e. a proper computer) before you can write a blog post. That’s a big barrier. With WordPress, all you need is a web browser or one of the various WordPress apps. I never found a way to update my Jekyll site from my Android phone; with Wordpress it is easy.

What I miss

I like WordPress, but there are some things I will miss. Markdown is great and much better than the WordPress editor or writing HTML by hand. I would love to see native Markdown in WordPress. There is also the small matter of speed. A static site hosted on GitHub is faster than a content managed site on cheap hosting could ever hope to be. And of course, WordPress brings its own idiosyncrasies and annoyances, the biggest one being its habit of tidying up your HTML whether you want it to or not. It’s still not the perfect system.

What next?

I switched to WordPress because it is still the best supported and best documented blog platform around. It works well enough. But I am always on the lookout for alternatives. I like the look of Ghost, particularly its Markdown support, but it is still a very new system for a production website. I am also interested in Editorially, which might be a good way of updating a WordPress site with content written in Markdown. The search continues.