Dmitrii 'Mamut' Dimandt

Can anyone explain the idiocy in some IDEs (VS Code, Xcode) with "temporary throwaway tabs". You open a file, it opens in a tab. You open the next file... the first file is thrown away and replaced with the new file.

Who in their right mind thought this is a good *default* behaviour?

What I love about this era of AI is the unlimited creative choices and potential for websites. Just look at this amazing diversity in website design over at @lovable_dev and @v0

If only there was some space left at the top in that horizontal bar at the top for the huge amount of 4 icons that we now have to cram into their own vertical toolbar on the left @Fastmail

Apple made a 24k gold and glass statue for Donald Trump Another tactic from Tim Cook to charm the president.

All billionaires are both cowards *and* enjoy licking the boots of wannabe dictators.

Apple in the EU: we're gonna fight your rather reasonable regulations with all the power we can muster

Apple in the US: www.theverge.com/news/737757/...

MacOS is worse than Linux now.

Updated from .6 to .7, and it stopped recognising the external monitor it's connected to.

(Also, of course the display I connected my laptop to is "extended". Who would *ever* think that a huge external 4K monitor could be main by default?)

This is @Truecaller. While genuinely a great call screening service, somehow they need to sell your data to hundreds of "partners" to do so

"You're in control" says @cursor_ai and disables the Continue button until you agree to give all data to them.

The actual environmental waste no one is talking about

[templates] A declarative JavaScript templating API · Issue #1069 · WICG/webcomponents Related to many template, template instantiation, and DOM parts discussions (but especially #777, #682, and #704) I wonder if we should add a JavaScript-based templating API? I think JS-based templ...

This strange hang up on "JSX doens't respect HTML semantics" (what does it even mean) and "our conclusion is that it must be template literals, so we'll fit all facts and narrative to fit the conclusion" is beyond bizarre to me.

And there's no answer who this is for: github.com/WICG/webcomp...

2/2

JSX is already compiled to optimised output in frameworks that use JSX. To optimised output for that framework, of course.

Or directly into DOM elements: www.solidjs.com/examples/cou...

1/

You know what would actually "increase efficiency, accelerate work and get context fast" @Jira? Actually fixing your dog shit of a software to be efficient and fast.

Hypermedia is a property of the client once you've unwrapped the contortionist back into their human shape, hypermedia is a property of the client.

Instead of writing a long twitter thread, I did the next worst thing: wrote a blog post.

Long, meandering, perhaps unfocused. Too late, it's out.

dmitriid.com/hypermedia-i...

TIL you can't edit bluesky posts to correct typos.

Oh, my attitude has been hones by over a decade of seeing web components people interact with others

And yes, JSX is a very loud confirmation of existing need. That web components people couldn't care less about

Here's Safari dev on this topic: x.com/rniwa_dev/st... and esp. x.com/rniwa_dev/st...

Justin is one of the better guys. As I said in a different comment, he is one of the few capable of looking outside the bubble. I even linked his posts.

I don't have all the answers, but I've seen the interactions. More importantly, I've seen *results*

And that's barely scratching the surface of all the discussions and all the issues.

Again, there are reasons why the vast majority of frameworks *including the most vocal proponents* just gave up on wcs beyond basic interop

3/3

Guess when the roadmap even appeared?

Hint: after most of the original people left Google, and new people formed a working group.

2/

Rich Harris on X: "So, web components • break a11y • break progressive enhancement (no SSR, broken without JS) • don't work with SVG • share a global namespace instead of being modular • ... Imagine how much tedious moralising we'd see if JS frameworks shipped with similar limitations" / X So, web components • break a11y • break progressive enhancement (no SSR, broken without JS) • don't work with SVG • share a global namespace instead of being modular • ... Imagine how much tedious moralising we'd see if JS frameworks shipped with similar limitations

Of course they weren't. Because the space was in flux. And web components *ignored every single voice that wasn't their own*.

Here's Rich Harris in 2019 listing just some issues: x.com/Rich_Harris/...

Guess when they were actually acknowledged as issues and when they appeared on wc roadmap?

1/

There's never been "collusion". It was a bunch of arrogant "we know better than you" 25-year olds who were given unconditional unconstrained uncontrolled reign over the web platform.

A decade later the sunk costs alone prevented them to accept any and all mistakes, and any and all criticism.

Nope. I'm not conflating. After years of ignoring, denigrating and gaslighting everyone else, now it's "omg JSX and declarative templating is good actually, and there are valid reasons why it is so popular"

It's ignoring it. There has never been vitriol on the part of these three. The amount of vitriol they experienced from web components? Oh boy.

(see rules about directives) etc.

So yeah. Nothing about lit is "more standard" than React.

2/2

It doesn't matter if it's "valid html". Template literals are literally wrappers for function calls (which you say is bad in JSX), lit-html has its own parser to parse and validate the syntax (like any other "non-conformant" parser), has its own rules about what can go where

1/

Framework authors not only wanted web components to succeed. They engaged at literally every turn.

The response? Ignore, deflect, degrade, gaslight. *For years*

4/4

And these are not the only issues, of course

There reasons why we he most vocal proponents of web components outside of the cult who even modeled their frameworks after them are now the biggest critics and want nothing to do with them: Ryan Carniato (Solid), Rich Harris (Svelte), Evan You (Vue)

3/

All the issues in this 2022 report (never updated since) were pointed out, multiple times by multiple people as early as 2016 w3c.github.io/webcomponent...

The reaction was invariably "you are wrong", "you don't understand", "this is a non issue", "use the platform" etc.

2/

Oh. It's been *extremely* bad. Including outright gaslighting directed at framework authors. Not to mention degrading and denigrating comments when people pointed out all the issues the sect begrudgingly finally admitted only after new people formed a "working group"

1/

The story of the past 10 years

PWAists: Apple cripples web apps on iOS!!!

Me: Android doesn't. Where are the good native-like web apps on Android?

PWAists: well, look at Twiter/X.

Me: really? REALLY? You call it a good web apps?

An irregular reminder that reddit still takes 2-3 seconds to load three times less text than old.reddit despite the backend returning all data in ~400ms. Oh, and a full page reload to load a comment subtree.

I think @jimsimmons.bsky.social should advocate bringing previous staff engineers back.

Justin Fagnani on X: "@zachleat A realistic plan to add something like JSX to the platform would be great actually." / X @zachleat A realistic plan to add something like JSX to the platform would be great actually.

And even before that, in 2022 x.com/justinfagnan...