This past weekend marked one month since I moved this site back to Hugo. It was definitely the right move.
This is actually the third time I’ve moved this site to Hugo. Historically, I get about a month in and start thinking about going back to WordPress due to workflow issues. This time is different.
Working on this Site Has Been Fun Again
First and foremost, working on this site has been fun again with Hugo. I’ve made over 300 commits since I started porting my old site to it. A few weeks ago I had even started learning Swift and playing with apps for Apple. I’ve enjoyed that, but I haven’t gotten far simply because I’ve had too much fun working on this site.
I’ve built the library and notes pages. I’ve optimized and re-optimized the site templates and I’ve built numerous scripts to make writing easier.
I’ve rethought the content, moving short, personal posts to the “notes” page in order to make my RSS feed and blog easier for folks to read. I’ve also consolidated the tags on this site, going from almost 300 to about 35.
So far I’ve largely kept the design, but that will be updated too. For now I’ve just been optimizing the heck out of what I have with more semantic markup and leaner CSS.
Through all of it, I’ve had fun with the tech. That’s something I hadn’t really had with this site in a very long time.
Workflow Matters
I wrote about my new Hugo workflow a few weeks ago. I’ve automated draft creation and publishing so that I don’t have to worry about the little details that used to drive me crazy.
Since that post I’ve taken it further by implementing the notes page. This page takes all the content that really didn’t deserve to be called a blog post and adds it to a dedicated notes section, separate from the normal blog (though still available via RSS). This corrects a long-standing problem: me publishing what was essentially garbage just to say I published something on a given day.
Going forward I’ve even built the notes feature into my publishing script. Any “Personal” post that is less than 80 words will automatically become a “note” upon publishing. In the end, this means the main feed will contain better content and I can easily publish the shorter, personal notes on the site.
In the end I’ve invested more in my workflow with this iteration of the site than I have with the site itself, and so far I’m really liking the results.
Performance and Privacy Matter
Finally, this site is far more performant and private than something built on WordPress and hosted on WordPress.com could ever be.
What started me on this path back away from WordPress was realizing just how bad WordPress had become for tracking. WordPress.com forces Jetpack onto sites and it was filled with scripts for user tracking, AI, and other garbage that I just didn’t need or want. While I had been able to remove some of that via code, it wasn’t enough. On top of that, no matter how much I tweaked the site the WordPress Dashboard was pathetically slow, especially on my iPad, and writing became a more frustrating chore than something I enjoyed.
Then there was the WordPress theme. I actually did like the base theme I was using, OllieWP (if I was ever forced to use WordPress again, this is the theme I would choose). In the end, though, it couldn’t overcome the shortcomings of working with WordPress’ full site editor’. For example, between the Dashboard performance problems and the ridiculous way editing actually works, it could take an hour or more for even the smallest changes. What’s worse is that there was no good way to roll anything back.
My final straw was trying to adjust the spacing between posts on the homepage. It took multiple screens, each requiring an independent save. In the end, I spent over 2 hours on it, hated the results, and then found that the only way I could roll it all back was to do a full database restore at the host level.
That’s just unacceptable.
Of course, then there were the front-end performance issues as well. Even though the page required no interactivity, I could never get it to load fewer than nearly a dozen JavaScript resources. Add in the CSS and other assets and the best I could get the page weight on the homepage was right around 500kb. That’s insane! Believe it or not, that was also an improvement from the 900kb it was before I started removing all the tracking scripts and other cruft.
Today the homepage of this site, with all resources loaded, is about 14kb. The content is the same as it was with WordPress but the bloat is just gone. This is good for everyone, especially those of y’all who don’t want to be tracked (there is nothing left for tracking other than logs my host, Cloudflare, might keep).
This Site Is Finally Something I’m Proud Of
In the end, after a month back on Hugo this site is finally something I’m proud of and can have fun with. It’s been a lot of years since I’ve been able to say that.
Building my own site is what originally attracted me to tech as a career. Compared to aviation, I was actually building something and it was amazing. The problem was that feeling of pride and wonder seemed to have died long ago.
Thankfully, neither was really dead. I just needed to think differently and spend some time focusing on what matters most to me: how I use my tools.