Do you have a source or benchmarks for the last bullet point?
I am skeptical that optimizations like that wouldn't already be implemented by postgres.
Edit: Btrfs has the worst performance for databases according to this benchmark.
https://www.dimoulis.net/posts/benchmark-of-postgresql-with-ext4-xfs-btrfs-zfs/
Although the usage of (x)wayland is novel, there have already beem projects which do something similar before.
Termux can run a linux container in a proot, which you can then connect to via an app like vnc to get graphics.
There exist several options to automate this setup, such as anlinux. There is also the proprietary andronix, which used to be open source but now it looks like tgere repos aren't being updated.
It's bad reporting to frame this as a novel app, when it's not. The novel thing is the way this app does xwayland rendered by a native wayland compositor (instead of remote desktop softeare or other solutions), which is really cool though.
As far as I know, there is no programmatic way to destroy an existing pizza. terraform destroy is implemented on the client side, by consuming the pizza.
I have a sticker of the nix one on a laptop.
have you looked at solutions which emulate github actions locally?
https://github.com/nektos/act this is one of them but I think I've seen one more.
Github actions also has self hosted runners: https://docs.github.com/en/actions/concepts/runners/self-hosted-runners
What would you use if you had a choice?
https://home.robusta.dev/blog/stop-using-cpu-limits
Okay, it's actually more complex than that. Because on self managed nodes, kubernetes is not the only thing that's running, so it can make sense to set limits for other non kubernetes workloads hosted on those nodes. And memory is a bit different from CPU. You will have to do some testing and YMMV but just keep the difference between requests and limits in mind.
But my suggestion would be to try to see if you can get away with only setting requests, or with setting high very high limits. See: https://kubernetes.io/docs/tasks/configure-pod-container/assign-memory-resource/#if-you-do-not-specify-a-memory-limit
In order for them not to be OOM Killed, you have to set the memory requests for them above their highest spike, which means most of the time they’re only using like 25% or so of their memory allocation.
Are you sure? Only limits should limit the total memory usage of a pod? Requests should happily let pods use more memory than the request size.
One thing I am curious about is if your pods actually need that much memory. I have heard (horror) stories, where people had an application in Kubernetes with a memory leak, so what they did instead of fixing the memory leak, was to just regularly kill pods and restart new ones that weren't leaking yet. :/
To answer your actual question about memory optimization, no. Even google still "wastes" memory by having requests and limits higher than what pods usually use. It is very difficult to prune and be ultra efficient. If an outage due to OOM costs more than paying for more resources would, then people just resort to the latter.
This is the same technology that lets people play windows games on android with good performance. Because there is not direct access to the GPU, they have to use GPU virtualization in order to get it access to a Linux proot that runs wine inside.
I'm excited to see it being used and developed in other areas.
design around ease of self-hosting. A non technical user must be able to self host easily and at a very low cost.
This may be a controversial opinion, but I actually like the way that hosting a lemmy instance is somewhat difficult to spin up. I like the way that it is requires a time investment and spammers can't simply spin up across different domain names. I like the way that problematic instances get defederated and spammers or other problematic individuals can't simply move domain names due to the way activitypub is tied to those.
In theory, you could set up something like digitalocean's droplets, where a user does one click to deploy an app like nextcloud or whatever. But I'm not really eager to see something like that.
Transferable user identity (between instances)
I dislike this for a similar reason, tbh. If someone gets banned, they should have to start over. Not get to instantly recreate and refederate all their content from a different instance.
Of course, ban evasion is always a thing. But what I like is that spammers or problematic individuals who had their content nuked are forced to start from scratch and spend time recreating it before they get banned again.
As for what I would really like to see, I would really love features that make lemmy work as a more powerful help forum. Like, on discourse if you make a post, it automatically searches for similar posts and shows them to you in order to avoid duplicate posts. Lemmy does something similar, but it appears to only be the title. It would also be cool to automatically show relevant wiki pages, or FAQ content, since one of the problems on reddit was that people wouldn't read the wiki or FAQ of help forums.
I would also like the ability to mark a comment on a post as an "answer", or something similar. I think stackoverflows model definitely had lots of issues with mods incorrectly marking things as duplicate, but I think it was a noble goal to try to ensure that questions were only asked once, and for them to accumulate into a repository of knowledge. For the all the complaints about it, stackoverflow is undeniably the one of the biggest and most useful repositories of knowledge.
There does exist a tool that does it. The creator posted about it on the fediverse. It only supported ubuntu at the time but looked extremely promising.
I cannot remember it's name. :/
Maybe it's linixify? But I remember seeing a post on lemmy with a youtube demo?
Me too but tapping the same spot to get to the same app still works so I just keep using it.