jeremy-code

$ vite build
[1/5] analyze client references...
vite v8.0.16 building rsc environment for production...
✓ 16 modules transformed.
✓ built in 57ms
[2/5] analyze server references...
vite v8.0.16 building ssr environment for production...
✓ 288 modules transformed.
✓ built in 379ms
[3/5] build rsc environment...

jeremy-code

TZ=UTC stat -f '%Sa' -t "%Y-%m-%dT%H:%M:%SZ" ./path/to/file.txt # last accessed
TZ=UTC stat -f '%Sm' -t "%Y-%m-%dT%H:%M:%SZ" ./path/to/file.txt # last modified
TZ=UTC stat -f '%SB' -t "%Y-%m-%dT%H:%M:%SZ" ./path/to/file.txt # created

jeremy-code

jeremy-code

react-diff-viewer-continued colors turned into tailwind css colors. I commented out the grays so you can use your own. Might want to remove black/white too

jeremy-code

import { useSyncExternalStore } from "react";

type Breakpoint = "sm" | "md" | "lg" | "xl" | "2xl";

type BreakpointWithUtility = Breakpoint | `max-${Breakpoint}`;

const useBreakpoint = (breakpoint: BreakpointWithUtility) => {
  const breakpointValue = window
    .getComputedStyle(document.body)
    .getPropertyValue(

jeremy-code

The Temporal API is very nearly baseline with support in 67.52% of browsers including beta opt-in support in Safari and support in Node.js 26. Of course, the polyfills @js-temporal/polyfill and temporal-polyfill also exist.

One of the larger points of contention is the parsing and formatting support in the Temporal API. Regarding parsing, in addition to object literals or epochNanoseconds/epochMilliseconds, there is currently support for parsing RFC 9557, JavaScript.toString(), ISO 8601, RFC 7231.[^1] While there originally was a proposal for a Temporal.parse API, it has since been abandoned based on the claim that "A type-span

jeremy-code

Switching from Radix UI to React Aria

Differences

No <ScrollArea>

Radix UI offers a custom <ScrollArea> implementation both as its own component and in some of its components (e.g. Select).

Per adobe/react-spectrum#7286, React Aria maintainers claim that due to accessibility issues and being unable to adhere to user system preferences, they don't intend on supporting a customizable scroll area. Per this tweet, it seems maintainer Devon Govett strongly dislikes overriding user preferences.

jeremy-code

import { Decimal } from "decimal.js";

const DEFAULT_MAX_ITERATIONS = 10;
const DEFAULT_TOLERANCE = 1e-10; // = 1/10_000_000_000

/**
 * Computes the continued fraction coefficients [a_0, a_1, ..., a_n] for a
 * **non-negative** finite number
 *
 * @see {@link https://en.wikipedia.org/wiki/Simple_continued_fraction}

jeremy-code

getImageDimensions TypeScript function using Image constructor and onLoad/onError. Much faster than Window.createImageBitmap. Uses imageOrientation = "from-image" to set width and height based on Exif data.

jeremy-code

Pdfimages seems to only extract images with transparency from PDFs by downloading it alongside the image mask. You can probably use something like ImageMagick if you only have a couple of images to convert it back into a PNG with transparency, but since there was a lot of images in my case, this is the script I used.

It uses pdf.js and sharp. Sharp was probably overkill but it makes the exporting part a bit cleaner.