I use IncusOS in my homelab. It's a joy to set up and use.

Migrated from Proxmox and manage all my VMs. Heavily use coding assistants to automatically set things up through the IncusOS CLI, translate Docker-Compose images to Incus, write bash scripts to automate launching new containers to use `--dangerously-skip-permissions` without fear of repercussions, etc.

What I love the most about it is that it's possible to manage IncusOS with declarative files, so you always have visibility into networking setups, resource configuration, etc.

Highly recommend checking IncusOS out if you have similar use cases!

As someone who has moved the opposite way, heavily using Incus and now checking out Proxmox, what made you go for IncusOS?

My gut feeling is that enterprise sentiment is leaning heavily towards Proxmox, fuelled by a VMware exodus that will only gain speed, and I don't see Incus really meeting the requirements most people have that previously used VMware, but of course Incus is awesome and you can't always pick technologies by what will be "employable" :-)

Having an OS that versions all my VMs and allows A/B running the actual OS, has a high quality CLI, is built on modern standards, supports declarative files, and it’s simpler than Proxmox.

I don’t really care for enterprise support. Incus hits a sweet spot no other solution does.

What solution do you use for declarative deployments? Last time I looked there was no default option?

I use my own solution (set of bash scripts) on top of IncusOS support for declarative files.

I haven't migrated to IncusOS just yet, but I use Incus on a server and have been pretty happy so far.

I personally like and use watchexec [1] to run commands on repeat.

Syntax example: `watchexec -r -e py uv run pytest -xvvs file.py`

[1]: https://github.com/watchexec/watchexec

Plus, I don't think Japanese eat a lot more fermented foods than other cultures. It's way more prevalent in South Korea, China, Russia, etc.

I agree with you. I followed him out of curiosity for one or two months, watched about 10 of his videos.

He seems to have a good intuition, but he gives weak and often cherry-picked reasonings, to the point that many of his takes are completely unreliable.

For a channel called Predictive History, he made too many weirdly precise explanations and predictions that turned out to be wrong. Then, he'd look over the old failed ones to find new ones.

That being said, I'd say his macro level analysis is directionally correct, as well as his read on the incentives of each party involved. Watch his lectures, but be skeptical and double check everything he says, because he does indeed make factual mistakes... some of them are caught in the comments by other viewers, some are not obvious.

I agree with your more thought out assessment of the channel.

I stil think the effort of trying to predict history (trying to understand causal patterns and extrapolate) is a valiant effort, and don't want to write his entire channel off. So I can't comment on his Iran take until I've seen it.

But yes, everyone makes mistakes and that just shows we don't have some universal theory for predicting history, so one shouldn't get obsessed with one school of thinking, but try it out, find its limits, and be open to other school of thoughts too.

He might indeed need some personal development himself. I followed him when the US bombed Iran's nuclear sites last year. He was involved in a controversy with his kid and turned into a dick, going into a charade against the Western education system, for being overly harsh to his kid in a public space and getting reprimanded for it. I'm not condoning his behavior, I wasn't there, but I'd take anything this guy has to say on personal development with a pinch of salt. By the way, he published a post of apologies in his Substack IIRC.

This is natural behavior as a father, specially from Asians. You do not know what actually happened to the kid, we too have done the same.

Have you looked into zmx? [0]

It doesn't have built-in notifications and there's no panel to see all the open sessions, but I wonder how hard that would be to add.

I've used zmx since I ran into it a few weeks ago. Uses libghostty as well. It's great because it allows me to replace tmux completely in all my ssh sessions, and can keep one session per assistant.

[0]: https://github.com/neurosnap/zmx

zmx solves persistence well, and I like their minimalism (not supporting windows, tabs, or splits). I think it's possible to make a CLI wrapper for zmx that adds notifications though, so you can have some niceties of cmux without switching to a new terminal. Lowkey we might explore this direction as well.

ive been working on glue for zmx+kitty (would do ghostty if it had proper ipc/scripting support). just changed the repo visibility on on gh cwelsys/kmux.

Does it do bufferbloat checks?

Is there any plan for this?

Funny enough, I actually just (2 weeks ago) added support for streaming from Pyspark to Polars/DuckDB/etc through Arrow PyCapsule. By streaming, I mean actually streaming, not collecting all data at once. It won't be released probably until May/June but it's there: https://github.com/apache/spark/commit/ecf179c3485ba8bac72af...

Not that I’m aware of. The Spark ecosystem seems a little too “stable” to be putting effort into that kind of development.

Edit: hah, based on the sibling comment, I stand corrected

I used it for close to a year and abandoned it because I kept running into issues with tabs getting randomly reloaded and extensions causing trouble.

What would you say has changed over the past few months? I just felt like Kagi wasn't prioritizing Orion development enough, being busy with their main Kagi subscription and all.

I'm not sure OP configured it this way, but it's possible. For example, yazelix gives you that functionality.

I do it using `nvim --server`, but not sure Helix can do it. https://neovim.io/doc/user/remote.html