Where main, roden, and margins are all their own hugo sites. They build into …/public_html and I have simple scripts to git sync and rm -rf public_html and rebuild all three.
I sort of walked myself into this setup in that last year I used /roden and /margins to test hugo out as /main was still Expression Engine. Finding Hugo to be pretty great, and understanding how it worked, I then set about converting ten years of essays and content into markdown, rebuilding the templates / theme, and clearing out a bunch of cruft that had built up over the years.
So now it’s all on Hugo, and I actually like the compartmentalization of sections into their own sites — makes testing easy, keeps directories clean, and makes updating just a small section simple as well.
A big thanks to all of you who have worked on Hugo, and especially @bep for building and continuing to work so diligently on this great platform.
Mainly I wanted to disconnect from any technical debt moving forward. I wanted my site to work no matter what — as long as Apache was alive, the site would be alive. PHP updates over the years had rendered bits and bobs of the EE site broken. It was too much technical debt for too little benefit.
Also, getting out of MYSQL was a bonus. There’s something nice about a bunch of frontmattered markdown.
And then performance. Although I’m running my site through cloudflare so it’s not that big of a deal, it’s nice to know it can’t get much simpler than it is with Hugo.
So while the old platform was doing its job, it felt increasingly like a duct taped Rube Goldberg machine, that could fail at any given moment. (Also, the EE install I was using was also 10 years old, not updated, and full of security holes I’m sure.)
Hugo’s active development (plus simplicity and speed) was the main selling point for choosing it.