Is it painful? I’ve upgraded several versions without difficulty. “third party” repositories are problematic so be sure you follow the guidelines about disabling them.

there’s always a bunch of people in the comments roasting them for AI use.

That happens here too, it’s just easier when you make AI disclosure a scarlet letter.

Well yeah, duh.

Opening ports is also trivial. You don’t generally need a dedicated IP if you keep a connection running but YMMV. And you don’t need a domain - just remember the IP. This is only for a few weeks.

Setting up ssh is nearly trivial. And saying it’s “less secure” doesn’t mean it’s insecure.

It’s the primary access point for thousands of systems and for good reason. It’s “good enough” for your little homelab.

Jesus… Just port forward 22 to your system and ssh in directly.

Just use ssh and be sure to disable remote root access and require key based authentication. If it’s good enough for aws, azure, etc. then it’s good enough for what you want.

You’ll want to read up on port forwarding and pick a remote gui tool. “No machine” works very well for x11, but rdp is supported by some wayland displays. Vnc is the old workhorse.

Why? It sounds fine to me. OpenSSH is the standard for secure remote access. And port forwarding works great.

OpenSSH is one of the most secure applications in the world. It’s fine to expose it to the Internet. Just be sure you have excellent passwords or don’t allow password auth.

Maybe I misunderstood - The “NixOS (bare metal)” seemed to imply to me that the containers were not bare-metal by omission. If not then my complaint is retracted.

I’ll die on this hill.

Containers run on “bare metal” in the same way that other processes on your system run.

I’ve been wrestling with this issue myself recently.

There can be both performance and other issues with a db whose data files are on nfs. That said I’ve done this for years without any real problem. No db corruption so far - but it’s a possibility and I’ve probably been lucky. I have good backups just in case.

Like you I also don’t want to maintain a “pet” database on my file server. I want to spin up databases in kubernetes pods and not use local pod storage.

I’ve been investigating using iSCSI instead. It’s “block level” rather than “file level” and performs better with a database. Also none of the locking issues mentioned. It’s a recommended way to do remote storage for a db.

The downside is that it’s less resilient to the file server becoming unavailable.

Might be worth investigating for you as well - I’ll be performing some poc tests to figure out how much of a pain it may be (I don’t want to need to fsck a bunch of targets after a file server reboot).

Is this useful? They’re started but doing absolutely nothing. Who cares? You need to plan for max usage not idle.

My forgejo instance running right now is using 1070MiB. That’s way off your 173MB.

You know what’s funny? Codeberg has said nothing about being concerned about liability.

Their terms service change simply mentions that they require certain licenses and that they have concern over the licensing of LLM meeting that standard.

Event their follow up communication says absolutely nothing of liability.

“It’s not…”

“There are probably law firms…”

This is wild unfounded speculation.

P2P torrent users are nothing like a code hosting platform. 🤣

As has been said - their liability for llm code is no different from hosting any other code. They already have the same risk. And that is typically that they must respond to take-down notices which they are already doing.

They are already dealing with all of the problems they would be dealing with with llm code.

It’s just code puritanism wrapped in pseudo legal justification. Own it.

This hasn’t happened. It’s a fantasy.

Maybe you don’t understand copyright?

Yeah - it’s well known that scripts to automate doing bad things did not exist before llms.

People have tried using forges to host explicitly illegal material all the time. Not just “this may be a copyright issue perhaps maybe”.

The copyright complaint is a fig leaf to hide their shame.

They also outright banned cryptocurrency code with no “copyright fig leaf” apologia.

It’s code puritanism plain and simple. “We don’t like these things so we’re banning them.”

People can upload code that violates copyright too. No llm required.

So if that is the concern have a rule about not violating copyright (which they may already have). And there is likely already a process for handling that.

This is the problem I have with the situation: the solution chosen (banning llm code) doesn’t really address the concerns they’ve raised directly. It’s just apologetics to make it sound like it’s reason-based rather than “we don’t like it”.

It’s their platform, they can do what they want, but let’s not pretend it’s not code-puritanism.