I don't think Scrum is right and our intuition is wrong. If the framework operates like a trick question that fools 75% of people, then the framework is wrong.
15.11.2025 20:20 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0@code-e-pendant.bsky.social
Full stack engineer, founder, rustacean, and WEEB.
I don't think Scrum is right and our intuition is wrong. If the framework operates like a trick question that fools 75% of people, then the framework is wrong.
15.11.2025 20:20 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Maybe, I don't know, I have a hard time believing the hype. Vibe coding doesn't even seem like a real thing to me
02.05.2025 02:25 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0Why is anyone even talking about vibe coding at all? There's nothing to talk about, this is weird
01.05.2025 01:06 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0A code snippet comparing the old way to set element properties, versus the new way. You used to have to select the element (whether from refs or otherwise) in the connected callback, then manually set the property like old-school JS. Now you can bind directly in the HTML templates using a special prop-colon syntax.
Finally added prop-binding syntax to Thunderous. I had my reservations because I think conflating props and attrs tends to cause a lot of confusion, but I think this compromise (inline with HTML, but distinct from attrs) adds a lot of value.
(Building a to-do app magnified the pains 100x.)
@mayank.co You deserve some credit here. You really helped me out with that Template component.
github.com/jonathandewi...
Next goal: A proper documentation site...
react-shadow-scope.vercel.app
It's live! React Shadow Scope now fully supports SSR and Form Controls!
Issues with layout shift and remote stylesheets have been fixed. 90% minimum test coverage for more stability.
www.npmjs.com/package/reac...
react-shadow-scope is getting a lot more downloads than I thought. Scrambling to release a new major version that fixes a lot of issues and key missing features like form controls
26.03.2025 13:38 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0Having been a huge Next.js fan for 5+ years, I have been over it for a while now. They get worse with every new update. It's to the point now that I consider it nearly unusable.
23.03.2025 16:55 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0AI is decent, just not nearly as great as executives claim. Also I want to not be harassed with surveys about it.
23.03.2025 12:20 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0No problem! And yeah that's not a jab at maintainers at all, it's just the nature of the beast, it can't be helped. But yeah, something like this definitely makes it easier to wrestle with.
21.03.2025 18:45 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0Granted, if everyone is just fixing things, you might get less feedback. Maybe, not really sure.
21.03.2025 18:41 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0I might reframe the question a bit - I find it more valuable to own the bugs so I can fix them myself. In one scenario, you open a GitHub issue and wait months for it to get picked up and worked into a stable release. In the other, you can just fix it in the same day.
21.03.2025 18:40 โ ๐ 2 ๐ 0 ๐ฌ 2 ๐ 0Well, personally I tend to just one-and-done generate them, then they belong to me to maintain from that point forward.
But there are ways though, like with shadcn: github.com/shadcn-ui/ui...
Hmm, I think maybe folks are complaining about the wrong thing. It's not that it's less maintainable, it's that it's more to maintain. IME, owning the code you use is almost always better.
Maybe this shift has yet to take hold in the WC community, but it's taking React by storm.
IMHO, the best UX of design systems is a CLI that generates components directly into the user's project.
See: ShadCN
I'm planning a similar approach but with WCs, but if you get to it first I can't complain
I'm finally back in the swing of things
www.thunderous.dev/docs/release...
Internet people say snide tactless things, but nothing OVERTLY insulting, baiting you into retaliating so it looks like YOU'RE the aggressor. Then they frame every defense as progressively crazier, until they rally a whole crowd to hate you.
Isn't the internet great?
I know I always talk about being a FC fan, but I do agree FCs should not be repeatedly called nor have stateful side effects.
But then again, no components should imho. Render once, bind directly to nodes, avoid complicated lifecycles and hooks.
CSS experts are very, very hard to find, I'm learning.
26.02.2025 00:52 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0Even in Sass, I've always felt that nesting encouraged bad patterns that result in specificity issues.
30.01.2025 14:28 โ ๐ 3 ๐ 0 ๐ฌ 1 ๐ 0Guess you did mention compiler and IDE tooling, haha. But now I wonder how far you can take TypeScript
30.01.2025 04:06 โ ๐ 3 ๐ 0 ๐ฌ 1 ๐ 0Type checking would have to be an IDE extension, wouldn't it? Not sure TypeScript is smart enough by itself, but I also haven't thought long about it.
30.01.2025 04:05 โ ๐ 3 ๐ 0 ๐ฌ 2 ๐ 0Is it more valuable for a dependency to be easily stripped out, or to be well-maintained forever?
15.12.2024 04:21 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0Ayy, remember this? Squashed!
www.thunderous.dev/docs/rendering
I feel like I can figure this out if I look close enough at the links already provided, but any chance you know where to keep tabs on this team's progress?
13.12.2024 21:48 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0While Thunderous may not LOOK native, its goal is to operate closer to bare metal without the complexity of a render lifecycle, non-native decorators, or a vDOM.
That being said, would it be a desirable feature to add some codegen to transform Thunderous components into plain web component classes?
Oh sick! I expected this to be one of those things that was too difficult to implement. Next question is whether we're going to see attention on it for the next several years... ๐ค๐ค
13.12.2024 21:13 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0Will there ever be a declarative version of adoptedStyleSheets? Is that even on the radar?
13.12.2024 20:46 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0Tbh though, I was never sure how I felt about this proposal. Breaking encapsulation for those who reject it? Is adoption worth more footguns?
Hmmmm.