An Experiment With My Website

The real problem is that programmers have spent far too much time worrying about efficiency in the wrong places and at the wrong times; premature optimization is the root of all evil (or at least most of it) in programming

So I’ve recently gotten the urge to tinker around with my website / blog (which really has only existed as a development playground up to this point, seeing as I’ve only ever written a handful of small posts across the years). The past several weeks I did a survey of the most popular website stacks — everything from the classic (and overwhelming at this point) Wordpress platform to the many nerd / developer-centric “markdown-to-static” site generators to newer contenders like Ghost.

These are perfectly acceptable solutions in their own ways, but honestly they all seem to have way more features than I need (vastly increasing the layers of complexity when there is something that I do want to tweak). Additionally, I’ve found that having a monolithic build system in place kind of shields me from what I as a software developer want — a holistic understanding of everything that is happening when I write and publish something online.

Of course, there is always going to be a certain level after which I no longer understand what’s happening. Whatever intricate webserver code Github Pages is using to serve this page and the CDN caching happening behind the scenes is far beyond what I can look into for right now. But as a primarily web-centric software guy, I do feel a strange sense of duty to be conscious of as many of the actual bits that I send up to the server as I reasonably can.

To that end, I’ve embarked on a journey for trentshell.com to start from the ground up and only add a file as it’s needed. As of this writing, I have an index.html file, this document I’ve writing now, a shared styles.css file, a CNAME file to configure the domain, and an atom.xml since an RSS subscription mechanism seems like table stakes for any website which has the ability to be updated.¹ New posts will be added in bare HTML. No templating, No build step, no Markdown conversion process. Of course, there may come a time when it makes sense to add those features, but that decision will be made when I have more history to evaluate the real time / efficiency tradeoff.

What do I hope to gain by typing angle brackets like an animal? A greater understanding of and attention to HTML best practices and standard-compliant markup, for one. A quicker deployment process if I go for months without touching this site again without needing to re-learn how to run whatever static site generator is being run under the hood. Mostly, though? All the serious street-cred that will come by me telling people that I update my website in the old-world, hand-crafted, artisanal manner.