Gone Away / Posts
Contact Followers

Milestone in ongoing blog migration

After months of working on-and-off on the transition to a new blog framework, I’ve finally hit an important milestone. Starting this morning, every post on Gone Away is being generated statically by Hugo — or at least it should be. The only page that should still be generated by WordPress is the home page. When I’ve finished figuring out exactly how it’ll look, I’ll migrate it too.

For me, the most fun and interesting new feature since my last update is the introduction of the journeys taxonomy. This is the centerpiece of the geocentric concept that’s been floating around my mind since even before I started my first WordPress blog in 2006.

An example of the journeys taxonomy, using my recent four-week Spanish course in Mexico City as an example.

An example of the journeys taxonomy, using my recent four-week Spanish course in Mexico City as an example.

The idea is that a typical blog post of mine is a narrative describing some intersection of time, places, and images. The conventional reverse-chronological order of blog posts captures the element of time well, but the element of place poorly. On the other hand, maps do a much better job of representing place. The journeys taxonomy will use both maps and chronology to tell a story. When you scroll the map and click a pushpin, you may also see an image, bringing all the elements together.

From the perspective of content, the big mountain I’ve had to climb is that I’ve only rarely attached good map coordinates to my posts. I’ve been working diligently to geo-tag my old posts, but it’s a time-consuming process. At the moment, only about 27% of posts have coordinates. It’s not impressive, although it’s a big improvement over where I was a week ago.

Concerning the technology, the migration is still a work in progress, especially when it comes to automating a lot of the processes behind the scenes that are still quite manual or semi-automatic. For example, as I’m writing this post, I’m saving it to a file, after which I’ll run a command to rebuild the website, after which I’ll run another command to upload the website to its host server — which, for right now, is the same EC2 instance on Amazon Web Services that I was already using for WordPress. I resized the image file locally and uploaded it separately to an AWS S3 bucket. Obviously the process can be made more seamless.

So, dear readers, please have a look around and let me know what you think — either by creating a comment on the blog itself or on its Facebook page. I know there are at least a few features that aren’t yet working quite correctly. For example, the maps are hard to scroll and click on mobile devices. I’m still figuring out how to make it work better.