Once upon a time, there was a blog.

This blog’s name was…well, I don’t really remember.

It didn’t have many readers.

It didn’t have much content to read, to be fair.

It lived at blog.trevor.io, but really wanted to be at trevor.io/blog. That’s where all the other content got to be. Why not blog too?

This blog wanted more. Knew it could bemore.

It just needed a break-out moment.

And so it waited…

It watched…




Every day, getting closer to the day it would fulfull its destiny and become the next BIG…


Hey folks! In this sprint, we threw away our old Ghost blog at blog.trevor.io (#sorry) and created a brand new one at trevor.io/blog 🙌😃.

Trevor.io blog image

Writing about what we do, and how we do it, has been on our list for ages.

And you know what they say. There’s never a good time to have a blog. You just have to get on with it and try your best to make it work.

In this sprint, our goals were to:

  1. Create a gorgeous blog / blogging environment (to make creating content as easy as possible).
  2. Have this blog at trevor.io/blog (not blog.trevor.io, which is generally not as good from an SEO standpoint).

Easy, right?!

Actually, not entirely.

This was a weirdly complicated sprint.

Long story short, Trevor.io is hosted on Heroku. Heroku has what’s called an “ephemeral” filesystem. That means that any changes to the filesystem are thrown away whenever your Heroku app (dyno) is restarted (at least once a day). WordPress, and its plugin infrastructure, is designed around storing most of its config as files on the filesystem, which makes hosting a wordpress blog a bit complicated.

Next step: we thought we’d just host the blog somewhere else. Using Cloudflare, we could take all the traffic directed at /blog and send it to a different IP address. However, when we did this, WordPress itself got confused and we were no longer able to login to WordPress as admins.

Now, it’s possible this was entirely fixable. However, we’d already spent longer than we wanted on this issue.

We decided to move our platform to app.trevor.io (still hosted on Heroku) and move all our landing pages to a new, more traditional hosting provider. We’d been thinking about doing this for a while anyway, so it seemed like a logical next step and a good opportunity to kill two birds with a single stone.

So, we opened a new account with Siteground (who have great reviews) and started moving our landing pages.

Most of our landing pages are built using Thymeleaf, the de-facto server-side (HTML) template engine for Spring MVC (for an excellent post on setting this up (recommended by the creator of Thymeleaf), check out Tom’s medium post from 2016).

To move to a new non-java hosting account, we needed to change this.

We decided the fastest way would be to use Jekyll, which is “a static site generator. You give it text written in your favorite markup language and it uses layouts to create a static website“.

With that done, we just needed to create the blog itself using WordPress.

We chose a theme we liked (surprisingly hard!), made some modifications to it, and added some plugins.

It’s a work in progress, for sure.

But we’re happy with it: it’s certainly a nice place to write, which is really the main thing we wanted to achieve in this sprint.

Let us know what you think. And don’t forget to subscribe! 🏆

Disclaimer: I wrote this post very quickly, several weeks after the sprint finished (having forgotten most of the details). It’s no excuse, but like everyone else, our schedules have been impacted by current events, so I was a little behind. I don’t plan on making this mistake again 🏋️‍♀️

P.S. Did you read our newest piece? 6 Metabase Alternatives You Don’t Need a Data Team to Use

Subscribe to stay updated

Our infrequent updates will make you happy 🙂 View our privacy policy

Comments to: Sprint 10: creating a blog we WANT to write.

Leave a Reply

Subscribe to stay updatedOur infrequent emails will make you happy 🙂