Cut the crap. Flatpak uses hardlink from repo where file names are jash of the file itself. The chance of duplication is exactly same as that of duplicate files of same name in same directory.

Flatpak repo grows because we trade uncertainty over abi stability with installing all needed versions of libraries. For abi incompatible builds you could already do that in many distros (versioned soname) but to a lesser extent.

Also I usually do not install nvidia GL with flatpaks that I won’t run on nvidia on hybrid gpu laptops anyway for energy reasons.

The real reason you won’t need antivirus.

<3

I do not get what you’re trying to say here, sorry.

On the note on similarity I mean macos userland is closer to bsd than linux. Also for normal usage freebsd is much different in nature than usual linux (free)desktop though they share same desktop shells which isn’t the case for macos either. And while most people aren’t writing with posix api everyday, many (most?) paradigms translate to win32 so that the crt from mingw works well. It matters only if you’re working with msvc toolchain, and then you’ve to adapt to windows-isms.

Personal anecdote: yes I find macos more familiar than windows even though I use windows vm often and macos rarely. At least the command names are same/similar… So your point stands, my point is more on the Aktually side.

Wait what? The default kernel doesn’t have a fuse fs, inbuilt or as kext? Didn’t know that. I thought all modern un*ces come with fuse.

Edit: It seems apple is introducing something called LiveFS similar to (but incompatible) fuse. Couldn’t find much docs and I’m not gonna read xnu sources rn.

underpants of MacOS is very similar to linux

no it’s not. xnu is very different from linux, with even design philosophy far apart. The userland (and bsd interface aka positive syscall world) is similar to *bsd’s, not typical linux userland. Only real similarity is launchd because systemd drew inspiration from it.

Isn’t is fuse? Why then it doesn’t work on darwin?

Only cute ones :3

For me it is the malwares. Other platforms do not stand a chance against windows.

Yeah I only find this which is bit technical. Anything else seems marketing bs. Seems like they’re making something similar to fuchsia by google but with linux abi compatibility.

Having linux shim alone makes it effectively monolithic like xnu.They even claim this by saying linux shim will hold global state in this otherwise microkernel.

Do you know where to find the HongMeng kernel? I couldn’t find in OpenHarmony gitee.

Proxy account or donor is outside US, in a suitable country

Edit: There is legal protection, so if that donor doesn’t have any potentially problematic business it’s safe ig.

My ex once got rickrolled in a public bus. I said the song is cool and kept it playing on speakerphone. The whole bus genuinely enjoyed it (people singing along), nobody complained :)

Well eating the fruit from tree of knowledge was a pretty big sin…

How to fo that?

questionable?

OpenVMS: ‘I’m in trouble’

most honest privacy policy.

No reason, few decades ago oven used to work just as well as they do today with knobs, thermostats and spring timers.

That’s why I said good question.

The oven I mentioned isn’t this smart but there exist ovens like

COOKING MADE SMARTER WITH WIFI POWERED BY SMART HQ: Voice-enabled cooking allows you to turn microwave on and off, add time or change power level via Alexa or Google Assistant; Scan-To-Cook Technology saves time and optimizes frozen food preparation

good question :)

I think it’s integrated ram inside the microcontroller. It stores states and programming (time, temperature etc) + the working memory for the program running on cpu. Surely some registers can do that but who cares.

I’ve an oven which when turned off in hot state while in convection mode will turn on the fans for few minutes next time I turn it on, regardless of mode and temperature. To overcome this bug I need to put mains power off for couple of minutes and let the caps keeping the ram alive drain. Not only it has hot state reset bug but also a ram initialization issue as well it seems. Thankfully that state is not stored in nvram.

The manufacturer was as expected: ‘we’re not software guy, we can send an ‘expert’ engineer (who knows only to replace parts, no debugging) and it’ll cost $$’. I thought I’ll reverse it and fixing someday, till then I’ll live with it.