On the blog

2026.05.21

This thread is now dedicated to discussing the blog system and further developments with the website. I have already stated above what I want from this blog but I have not gone into detail on how I plan to do this. I see a few options.

Fzf and tags are dynamic...

The first feature that causes contention would be search. I both want a tag system and fzf. Both of these require* JavaScript. But I believe static-first architecture helps at avoiding entropic debt so I use JS only where interactivity genuinely benefits the UX. This also follows for any dependencies. The more you have, the more points of failure and more importantly more places that stack complexity.

This is partly why I also use so much abbrs so much, didn't require a single line of JS, the only thing I did was add underlining: 
abbr[title] {
  text-decoration: underline dotted;
  text-decoration-thickness: 2px;
  text-decoration-color: 
    color-mix(in srgb, currentColor 50%, transparent);
  cursor: help;
}
Even this ^^^ code block right here requires JS for copying to clipboard. The reality is that JS just cannot be avoided without being really boring.
The other reason is just UX. It works just like parenthesis but without having to bloat up the screen for those who already understand. This is heavily inspired by Ted Nelson's take on linkage.

Regardless, I have implemented fzf before, specifically in the search bar of the Resource Center of which that link is just one design I made for a local non-profit. The solution I used was very inline with Jamstack by just linking to a third party js file https://cdn.jsdelivr.net/npm/fuse.js@6.6.2 The only real issue I have with this is that given it's not an API, I should have just been able to downloaded the file myself and mirror it instead of creating another dependency. So I wont make that mistake this time.

My plan

A question I still have to answer is how to store everything. I think my conclusion is to have a RSS feed that links to all the different posts, to have them as separate files.

The blog.html page will be generated to list out all the posts and will have a way to narrow down which you want to view by a bunch of tags you can toggle and a fzf search. By the time I push the commit with this page I will probably have already done that.

I have been given a super cool opprotunity to code C for a project in the fall and it's tempting to take every project and see if I can do it in C, but in this case Imma play it safe with something quick and easy. I am writing the static generator of the blog from RSS in rust. afterall, it's one of the main things I want to advertise myself as competent in and it's been a while since I actually used it.