dieulot

No dependency on GNU required anymore, POSIX 2024 supports it: https://blog.toast.cafe/posix2024-xcu#the-null-option

It was made a release candidate in the previous (non-LTS) major version, in v25.7.0 from February.

Release notes: https://nodejs.org/en/blog/release/v25.7.0

Issue: https://github.com/nodejs/node/pull/61262

I'm wondering if the Xcode simulator (without Xcode running) performs as well, my 2020 Intel MacBook Air has been incapable of running Safari in iOS smoothly for nearly all its life.

Charts: https://sixcolors.com/post/2026/04/apple-announces-record-fi...

https://archive.is/8rPUA

That’s an unreliable way of simulating an unreliable network, as overviewed in https://calendar.perfplanet.com/2016/testing-with-realistic-...

There’s no statement or action (such as banning menu-bar-only apps from the Store or even changing the APIs) supporting that Apple still wants menu bar items to be ephemeral.

Could you back up your assertion that Willy Tarreau — who used to maintain the Linux kernel — is “an AI enthusiast”? I can’t find anything about it.

No, it’s >1 second on every machine.

Regarding HAProxy, they ended up using AWS-LC in their new Debian/Ubuntu “performance” packages: https://www.haproxy.com/blog/fresh-from-aws-reinvent-superch...

The issue of rootless malicious command overrides is solved by typing the whole path, such as "/bin/sudo".

(2015) https://web.archive.org/web/20150115000000*/https://www.howa...

Certbot does too as of 4.0.0 (2025-04-08).

This is due to one of your AAAA records: ::ffff:157.245.83.16. (The other one is fine.)

CLI:

    dig aaaa saile.it
    curl -v --resolve "saile.it:443:[2a04:4e42:2::775]" "https://saile.it/1145-pull-requests-per-day/"
    curl -v --resolve "saile.it:443:[::ffff:157.245.83.16]" "https://saile.it/1145-pull-requests-per-day/"

- https://www.ssllabs.com/ssltest/analyze.html?d=saile.it

- https://toolbox.googleapps.com/apps/dig/

To be more explicit: asking hn@ycombinator.com why some submission vanished in a mysterious way does bring a clear answer in my experience.

HTTP/2 “pipelining” (multiplexing) is still there and works as intended; but bundling is still much more efficient.

This article delves into just that: https://csswizardry.com/2023/10/the-three-c-concatenate-comp... The first pair of waterfall graphics illustrates the problem clearly.

(What you vaguely remember as not working as intended is probably Server Push).

Craig Federighi […] merging iOS and macOS team together.

What year was that?

https://archive.is/U93VS

No because that additional CSS file can only be fetched after the initial HTML file has been fetched, not simultaneously; thus requiring another roundtrip to the server. (HTTP/2 Server Push promised otherwise but it was never implemented correctly.)

Latency being much more constraining than bandwidth, embedding a stylesheet that’s small is nearly always the correct tradeoff.

“<file foo” is POSIX: https://pubs.opengroup.org/onlinepubs/9799919799/utilities/V... (see the absence of mention that redirection needs to be after the command)

What would a specific table of contents element bring over a <nav> inside your <main>/<article>?

(As demonstrated in the first example of https://html.spec.whatwg.org/multipage/sections.html#the-nav...)

Where is said track record of the main author being an ass to people who aren't willing to be marginalized, or anything along those lines?

If we become too insular, then we doom our larger economy to the same mistakes France, Germany, UK, and Japan made.

What are those? (Besides Japan’s which you’ve answered here https://news.ycombinator.com/item?id=40143090)

You can `set +e` before and `set -e` after every such command. I indent those commands to make it look like a block and to make sure setting errexit again isn’t forgotten.

For a more complete view, the HTTP Archive tracks general site weight over time and gives you percentiles (click Show table): https://httparchive.org/reports/page-weight#bytesJs

The median JS bundle is 600kB on desktop. p90 (“high 10%”) is 1830kB.

It’s an additional abstraction.

On Windows/Linux closing an app requires doing the obvious thing. Press the big X, you’re done. You’re back to the previous app instantly.

On Mac after closing a window what you see on the screen is a lie (besides the ever subtle menu bar and shades of gray); you’re still in the visually-hidden app.

For GFW reasons, the Chinese users will fail to connect to github.

Since when has GitHub been blocked in China? I thought the government reluctantly let it pass.

America means “french-door” as in https://fr.wiktionary.org/wiki/French_door

That piece of news is about only a week of pause for a version that’s even buggier than usual. Toss your Snow Leopard dreams.

“Flambéur” isn’t right but “flambeur” means something else entirely (gambler). “Flambéer” would be more legitimate but would be hard to decipher. The title tag plays it safe and uses “flambé restaurant server”.