They’re trying to exploit what they consider a loophole in the AGPL to stop people from forking the project.

At the same time they refuse to accept any PRs.

FSF has a great write up about it.

https://www.fsf.org/blogs/licensing/agpl-is-not-a-tool-for-taking-freedom-away

Strong disagree. China is a reliable if not self interested partner.

The US goes from selling you weapons to threatening to invade you randomly at the flip of narcissistic coin.

You’re allowed to not fiddle with it.

I can see how for some people cron is more straightforward to learn, at least till you need to handle logging, checking for cron results, handling when the triggered event can’t happen that instance, ensuring only one instance of the triggered thing happens at once, adding time jitter, etc.

Then timers are way simpler. Timers let you create robust timed events for free. With cron you need to do all that yourself.

That’s because you know cron. If you knew timers equally as well they would be easier. And they let you handle the edge cases (retry, randomness, tracking, logs etc) without the need for a custom script.

Once you factor in the production edge cases I think timers are clearly easier. You get all of it for free.

I think if you know cron from the start it can be easier, but it gets really annoying really fast.

Compare:

0 0 * * * /usr/bin/flock -n /tmp/myjob.lock bash -c 'sleep $((RANDOM % 3600)) && /usr/local/bin/myjob.sh'

To:

[Timer]
OnCalendar=daily
RandomizedDelaySec=1h

That and things like systemd preventing overlapped delays, handing what to do if the system was down during the last cycle, built in logging and event tracking. Seeing successful vs non successful runs etc.

Once you add in those production requirements cron gets annoying fast and timers are easy.

The main functional difference between systemd and others is that systemd will just work. Others will require you hand tune and hand tinker with a non-mainstream Linux distro.

If your hobby is init systems by all means mess around though.

I personally quite like systemd. Unit files are clean, timers services and sockets are easy to manage etc.

Honestly it’s a non-problem. Best advice is to use what is best supported. Don’t let the extremely fringe (but loud) tiny group of systemd haters throw you off.

Fedora with Niri on laptop / desktop.

Alma Linux on servers.

Out of curiosity what mandatory features would you consider missing?

More up to date news. It has 12GB of vram and competes with the 4060.

Should be available to buy June 18th.

https://www.tomshardware.com/pc-components/gpus/china-firm-lisuans-homegrown-6nm-g100-series-gpus-announced-with-up-to-12gb-of-vram-lx-7g106-can-play-cyberpunk-2077-and-other-popular-steam-games-arrives-june-18-in-china

I think you’re mixing things up here. Mac’s are actually VERY performant, both objectively and relative to the on board ram.

There are many critiques that can be made of macOS. But saying it runs inefficiently is not one of them.

Maybe in 2019. In 2026 8Gb of ram alone costs $150

The same applies to this guy. His work is quite impressive, but his antisocial tendencies got him booted from working in the kernel. And now he’s gone down this path.

Hopefully things level out before we see a templeOS level mental health crisis situation.

Feels very Telekom to do research on using existing wires. It’s like their thing.

Look at the Jellyfin one. Why is there something breaking the logo from the right? Lots of little things.

Fair. It’s not just that one though. I notice a lot of weird things there.

But I responded to a comment cursing out systemd on a post about system V being dropped?

The comparison is as made because the comment brought up systemd on a post about system V.

Why post this ai generated content? Since when does the docker logo have the Cassandra eye in it?

Yah! Screw them 20 line unit files. We roll with 500 line bash scripts.

/sarcasm

I mean yah? That’s obviously true. That’s why I said it was close to actual anti semitism. Not bona fide antisemitism.