This looks cool, I always wanted way to represent the result of ± in data of simulating stuff like the quadratic equation. victorpoughon.github.io/interval-cal...
Posts by Angus Khan
When I'm procastinating I find I have a bad habit of picking up of tasks with an infinite illusion of progression, it feels good to see things get done and they can go indefinitely. But it's just an illusion and an unhelpful distraction.
I also plan to make shadow dom a primitive in the rendering DSL, along with providing mechanisms to better handle shared styles sheets between shadow doms.
Along with some level of support for web components and custom elements, main headache is typing atm.
The goal with my client side rendering library is to write a property UI without direct interaction with the DOM APIs, allowing me to render stuff in a web worker. Main headache is stuff like active event listeners, and focus handling, although I have some ideas.
I would like to rewrite it tho, so that you just have directives at the top of the page that specifies an elements template location without repeatedly.
But if my client side rendering library gets good enough, I might just be able to render the HTML via that (so I will need a build system).
Main goal here is to remove layout jank/shift on load.
While ensuring I can continue to work on it without a build step (so I don't become a prisoner to it). While providing a reasonable developer experience.
I added some templating logic to the bundler I made for akst.io/notes
I wrote it in a way that it'll still work in the event I open the HTML file in the browser as is without a build step, but I setup the dev server to interpret the directives to simulate the shipped experience.
It doesn’t really matter what app I’m using, if I type it on the iPhone keyboard, it’s gonna be fucked up in special unique way and I often miss it as it often happens while i’m in the middle of something else.
What the hell happened to iOS
Has anyone else's spotlight search on macOS just kind of stopped working in the latest major version of MacOS?
Bass Tabs for Hope Tala's DTM
Felt sad to see this
They’re looking more perky today
Approximating log(x) with a Taylor series of degree 1 through to 100
As they say “he has rizzen”
My Claude bill has skyrocketed in the last month, my usage hadn't changed all that much, but I was getting top ups multiple times a day over once a week.
I ended up switching from iTerm2 to Ghostty. I realised there was a bunch of mouse features in neovim like right click goto definition that existed this entire time and I didn't know
The Australian free public transport debate on twitter would probably more constructive and effective with promoting their ideas if their main proponents weren't so dogmatic and aggressive towards anyone who doesn't immediately doesn't agree with them.
Okay, Kagi's April fools is kind of cool, that said this just my website all year around
Maybe this is why they call it green square
Everyone in green square is Irish this morning it seems.
The whole bsky and X fragmentation is tragic because some of the people who left X (like the entire react term) are not involved in the whole pretext discourse
Would love for better platform support this kind of text layout as well.
x.com/_chenglou/st...
I had a similar experience delving into to earlier RBA reports between the 1960-1980s. Whoever's job for preparing graphics clearly took a lot of pride in doing so.
Like there's stuff that are just artefacts of the medium (like the way colours mix in print) that you don't see these days.
It's kind of frustrating how few general tools exist to solve math problems like deriving limits and derivatives despite being largely deterministic.
Like such tools exist, but to do it yourself you end up having to write a lot of code.
Lore accurate delin
Added delin from shenmue to my island in tomodatchi youtu.be/qEWfxS0lMjs?...
Okay I figured out how to actually show an actual plot in the terminal
The problem is though using exp on the will almost certainly result in an overflow.
But if you're dividing by another equally large logged factorial, it should be fine to use exp on the result. Such a case being something like n choose k
I've just realised you can use log's to approximate large factorials to avoid overflow.