Rewriting my site... again
This is the fourth time I have started to rewrite my website in equally many years. Let me explain why.
Why I am writing my own site
When I made my first web page, in 2020, I did not have any experience with web development. For that reason, I simply chose an open source template that I liked the design of and tweaked it. 1 However, I wanted to change more than just the text or hero image on that page. I also wanted to change the design of some sections, re-arrange others, and ensure it was functional without Javascript. This proved difficult, as I now had to navigate large HTML files that were primarily styled by classes from style sheets large enough to warrant their own table of content, using pseudo-classes and media queries, and backed up by various Javascript functionality, all while never haven written a basic website before. 2 While in the end I managed to make the changes I wanted, it often took longer than I wanted to, and I did not really feel like I learned much in the process.
So when I came back to update my page 2 years later, I knew I had to make a change. I was going to build my own site. A site I would understand, could easily modify, and, just as important, allow me to build a gradual understanding of the basic building blocks of the web. The result was a placeholder website that looked worse and took me longer to build than adjusting the template would have. But I understood every bit of it, the site was fast(er), and building it was fun.
Replacing my placeholder
The major problem with projects is that they are much easier to start than to finish. The "first" placeholder page I wrote stayed up for a whole year. When I did finally get around to replacing it, I at least finished the frontpage and the 'about' page. I did use some AI art which was intended to be a placeholder but, you guessed it, I never got around to replacing it. It also had a really annoying visual bug I could not figure out. But I did improve my understanding of HTML/CSS, as well as optimizations like subsetting fonts, preloading, and image compression. All in all, a fairly good local optimum to let things be for a while.
The page that never was
Another year or so passed and I had learned CSS animations exist. I had already considered getting rid of my front page images altogether, both for optimization and because I did not want to use AI images any longer, so I figured I replace them with animations. To keep a long story short, I took too big a bite and couldn't finish my project in the time I had for it. I decided to cut my losses and get back to contributing to OpenML and the AutoML benchmark. The site never went live, but at least I did write my first CSS animations and played around with the GitHub GraphQL API.
My latest website
And now you are reading this post on the site I wrote when I got yet another itch. This time I was a little more deliberate though: I kept the scope small, and largely confined to concepts I was already familiar with. But I also knew I will want to keep experimenting with things in the future. For that reason I made sure include a blog section. The next time I have an itch, I'll write a single page, it can be a simple demo, and it doesn't have to be a multi-page responsive website.