Mastodon

The Ruby ecosystem is going through some pretty hefty turmoil in an attempt to rescue and restore the integrity of their community.

I’m even seeing some good efforts in PHP-land to better distance itself from the vacuum of stable leadership in WordPress too.

Is there any such movement in JavaScript (the same problems exist)? The biggest pattern I’m seeing is a *huge* centralization in project funding!

The absolute UX *drama* of the new tinted (vs. clear) Liquid Glass preference in the macOS 26.1 System Settings. Why separate from the Reduce Transparency accessibility toggle? Why so prominent in the Appearance section? Why do users need to know about Liquid Glass at all? Mistakes were made here 😅

as “industry” survey season rolls around again it’s important to remember that the results from these things reflect the biases of the most popular dev communities that promote them (and then we get to all pretend like the results are meritocratic)

I now believe it’s better to not participate at all

Thinking about dark/light mode pickers that have three options: Auto/Dark/Light.

If the default is “Auto” — in what scenario might someone want to move away from Auto (which would visually be a noop to the same theme?)

I’m using Light mode as my system preference, but it seems unlikely that I would swap from Auto (Light mode) to Light mode (not Auto) on a specific site

“AirTag Found Moving With You”

The ominous alert. Fight or flight — adrenaline. After surveilling my surroundings I was frightened to discover that AirTags weren’t the only things moving with me. There were multiple *people* shadowing my movements as well. I was on a bus.

TIL that CSS attr() cannot be used to compute anything in `url()` as the spec considers these to be “attr()-tainted” (for very valid security reasons): https://drafts.csswg.org/css-values-5/#attr-security

Thought maybe there was a way forward for global custom CSS-only favicons via new advancements in attr() https://developer.chrome.com/blog/advanced-attr (but alas, no — which, in hindsight — seems valid and fair)

is there really no low-level mechanism to intercept boolean primitive coercion in JavaScript?

why can’t I extend `Boolean` to make `new MyBoolean(false)` be `falsy`? (I understand why it isn’t, but I want to make it so)

> Note: Unlike other type conversions like string coercion or number coercion, boolean coercion does not attempt to convert objects to primitives by calling user methods.

https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Boolean#boolean_coercion

folks migrating away from hosting providers that hadn’t previously considered how tightly coupled their web framework was to said host are about to learn some things — I’m sorry and I wish you the best on your journey

@ricmac do not look this up but it is still live on his web site. He posted naked pictures of his family (including his underage kids? wtf??) on his web site as a “naked christmas card” — surprised he didn’t get in more trouble for it tbh. For as much conversation that goes into influencers posting stuff of their kids before they know what consent is, it’s wild to look back on this

Crashing out at how poorly npm (Microsoft) is handling this security incident. Eleventy is not affected any more but *lots* of other tools in the JavaScript ecosystem are!

Hours later and the compromised package versions are still public…

Maybe don’t install anything from npm today, folks.

Edit: the issue was resolved at approximately 18:30Z