Terminal? Hard to be sure, but I think there are glimmers of hope that the answer is "no" in the short-term.
Corollary question: should it be? Eg, is "liberal democracy" really the best we can possibly do? My take is that the long-term goal should be a society based on Voluntaryism with no use of force for anything other than self-defense. But if we ever get there, it won't be soon, and in the near-term the collapse of liberal democracy is trending towards the full-on advent of fascism and totalitarianism.
So at least for now, I believe liberal democracy is something worth fighting to protect.
It wasn't literally today, but about 2 days ago I discovered the -H (and --host) arguments to systemctl. These allow you to pass a hostname of a remote server, so you can use systemctl to manage systemd services on remote machines. It layers over ssh, so the cleanest way to do it is to have passwordless ssh with certificates set up to the remote host. If you do, running systemctl commands remotely is totally seamless.
Telefork reminds me of my days writing MPI code on Beowulf clusters! I even took a stab at a similar'ish idea, that was meant to make running code on distributed clusters in Java "just a library" that (somewhat) emulated the existing Threads API. I actually got it to work, but I don't think it was practically useful for anything. More of a learning project. Still, it was fun. :-)
For me personally, I'd say "not much". I'm used to using the latter form, and it's fine. But I will say that now that I know about this, I see using it in scenarios where I'm running a lot of systemctl commands over and over again in close proximity and have the "muscle memory" of typing "systemctl" more in mind.
For example, when working with a new service that isn't quite working right yet, and doing many iterations of:
Especially if I'm testing on my laptop AND a remote deployment, I think it's easier from a cognitive viewpoint to always "think systemctl" instead of having to "think systemctl" sometimes and "think ssh systemctl" in others.
To be fair though, it's all a pretty minor point. But I do think it's cool that systemctl has that option. shrug
But that seems worse? If all your commands are systemctl, then you just have to ^p and edit/rerun, but there isn't a `systemctl $EDITOR /some/file`, so you'll have to ssh to run emacs (or use TRAMP, if using emacs), at which point you might as well just have a shell on the other end and do everything without having to tack --host onto your commands
I don't know. Maybe. It seems like something that could be handy, but I don't yet know how much I'll wind up using it. For what it's worth, I was working on a scenario exactly like this earlier tonight, and I wound up just ssh'ing into a shell on the remote box to do stuff. But that might just be inertia from being used to doing that, or maybe it was because I had a lot of other stuff to do on that box at the same time...
Was wondering the same. "Do one thing and do it well" philosophy...
If I want to run remote commands, I know how to do that already (namely, as you've shown), why learn to use a new tool for it? It's like these options to tar for compression: if I want to have a gzipped tar I'll run the tar c /sdcard output through |gzip thank you very much! (and drop in zstd instead)
Sad news. Dilbert was a big part of my life for a long time, and brought much laughter and enjoyment to my life. But on the other hand, later in his life Scott said a lot of things I found frankly repugnant, and Dilbert more or less disappeared, all of which made me sad. But he was still an amazing writer of comedy at his best, and I hate to know that he has passed. Plus, every death is at tragedy for somebody - friends, family, loved-ones of all sorts - whether we specifically like someone or not.
Of course you can. It's an amd64 processor with a different name and a coprocessor. Linux doesn't have any problem installing on UEFI, so unless they did something weird, it shouldn't be a problem. Apparently linux runs pretty well on the AMD AI cpus (benchmarks of lunar lake primarily vs Ryzen AI 300 in OP):
But concept-wise I think it's a dud. A laptop without a monitor. So you either carry the laptop as two pieces (some good portable monitors around) or you are limited to only plugging it in where there's a monitor available.
Also, it seems they couldn't be bothered to spring for mechanical switches when the entire product is focused around a keyboard. Sheesh.
FWIW, I still use bash as well. Nothing against zsh per-se, it's just that I know bash, bash works, and there's no particular pain I experience using bash that will obviously be solved by switching. And when you factor in anticipated switching costs, I haven't found any compelling reason to spend any significant time on zsh so far.
I haven't straight up cancelled my ChatGPT subscription, but I find that I use Gemini about 95% of the time these days. I never bother with any of Anthropic's stuff, but as far as OpenAI models vs Gemini, they strike me as more or less equivalent.
Corollary question: should it be? Eg, is "liberal democracy" really the best we can possibly do? My take is that the long-term goal should be a society based on Voluntaryism with no use of force for anything other than self-defense. But if we ever get there, it won't be soon, and in the near-term the collapse of liberal democracy is trending towards the full-on advent of fascism and totalitarianism.
So at least for now, I believe liberal democracy is something worth fighting to protect.
reply