I guess I don’t understand. You followed the docker installation directions correctly and it didn’t work or you modified the directions in a way that you prefer and it didn’t work?

I have it installed for a few years now. I started with the AIO but moved to the separate container install after AIO was deprecated. I imagine the install process is too complex for portainer. https://docs.funkwhale.audio/stable/administrator/installation/docker.html

I did steps 1-4 and skipped the rest because I already have a proxy server running. Don’t remember anything related to snapd though. Mine is running in a Debian 11 VM on proxmox instead of an LXC, but the process should be the same. Also they have a matrix channel for help https://matrix.to/#/#funkwhale-support:matrix.org

From what I remember it was relatively painless to install, but upgrading can be a chore, especially this last upgrade. My main interest in FW was the federation aspect as far as finding new music. If you don’t care about federation, maybe a simpler option would work better for you.

Feel free to link us to these abusive communities then since you are making a singular claim yourself.

Yes I watched your video. So many wrong comments from ignorance. And blaming a protocol for the behavior of people? Another commenter here pointed out how Tor could be blamed in the same way. Why not just do away with the internet then? Ridiculous.

Nothing wrong with the matrix spec. I’ve ran a matrix homeserver for several years now and have never had issues with illegal material. It’s free, open source, and self-hostable. If you have an issue with matrix.org that’s a different issue. Don’t use matrix.org servers. In fact, they have made it a point in their literature that they want you to use your own servers, they preach against centralization, and point out that matrix.org is an entryway to get you started, not an endpoint.

At the very least you need to install a webserver and you need a proxy of some kind. If you truly want old school you can just create html pages hosted from the root of your webserver (although there are now easier modern ways to do this, you might learn more the classic way rather than using a CMS).

You will want a reverse proxy to lie between your webserver and the internet that handles SSL. Let’s Encrypt is a good option to generate a cert so that you only expose port 443 on your router to the internet and your webserver. You’ll have to open port 80 to generate the cert but can close it again once generated. Then you will have https.

That’s the basics. The how-to’s are easy to find online.

I’m not sure about the software always breaking, I haven’t had this issue. I will say though this most recent update (19.4) has me frustrated, mainly because the instructions are clear as mud (especially pictrs 0.5). Once I get it figured out I’ll have to post a real upgrade doc instead of what is currently available. I have never seen the lemmy matrix as busy as it is now with upgrade questions and puzzled admins.

I’m not sure how soon you need this, but if you can wait sipeed has a $20 kvm with ATX control that should be out soon https://lunar.computer/news/sipeed-announces-new-20-risc-v-kvm-device/

There’s an interesting book I read recently related to this called The Anxious Generation: how the Great Rewiring of Childhood is Causing an Epidemic of Mental Illness. I’d recommend it.

As a counterpoint, EFF put out this article today: The Surgeon General’s Fear-Mongering, Unconstitutional Effort to Label Social Media

This is the part of the new compose file that I don’t understand:

PICTRS__OLD_REPO__PATH=/mnt/sled-repo PICTRS__REPO__TYPE=postgres PICTRS__REPO__URL=postgres://user:password@host:5432/db

When I try to add it, pictrs complains about not being able to connect to postgres, even though the credentials are correct and I’ve already run the postgres migration script. Did you not do this part when you upgraded?

I am using Kinoite for quite a while now and not once did layering break anything.

That’s great for you. Not everyone may use their distro in the same way as you.

https://discussion.fedoraproject.org/t/is-silverblue-rpm-ostree-intended-to-be-used-with-layered-packages/26162/2 https://discussion.fedoraproject.org/t/fedora-silverblue-36-will-not-succesfully-deploy-after-layering-packages/77502/3 https://gitlab.gnome.org/GNOME/gnome-software/-/issues/991 https://github.com/coreos/rpm-ostree/issues/4280

Not to mention the whole Firefox debacle of including an outdated borked version based with the system install instead of just moving to Flatpak install of most recent stable release. There’s a very valid reason why package layering is discouraged by atomic maintainers and why toolbox is there by default as part of OS. And don’t even get me started on DKMS and driver installation.

So, the points in favor of Kinoite is sticking closer to upstream, however it seems like I would need to layer quite a few packages. My understanding is that this is discouraged in an rpm-ostree setup, particularly due to update time and possible mismatches with RPMFusion

It’s not only discouraged but often times it’s system breaking. I used Kinoite for a year before I just became too frustrated and gave up. The first thing I learned though was to stay away from package layering because it tended to break things more often than not. Basically if you can’t find or build a flatpak and you don’t want to use toolbox all the time, just stick with workstation. Immutable is great when deploying to multiple servers or locked-down corporate workstations, but it makes no sense for your personal setup especially if you’re already familiar with Linux.

Even more shocking is the revelation that somehow, even private DMs from Mastodon were mirrored on their public site and searchable. How this is even possible is beyond me, as DM’s are ostensibly only between two parties, and the message itself was sent from two hackers.town users.

I find this hard to believe but stranger things have happened.

Just keep in mind that even with a jetson board you’ll need one of the higher memory configurations to have a non-frustrating stable diffusion experience. 32-64GB like the Orin and those aren’t cheap. The nanos just don’t cut it without severe optimizations and very long generate times.

Elaborate on why samba is bad when it comes to security? Like list a bunch of links like this or write a paragraph summarizing them like a chatbot?

NFS does symlinks but they have to be configured correctly.

Samba may have not given you issues in the past, but it also doesn’t give you any security.

Also never use samba. At the very least use NFS.

That’s too bad. Nova’s usually a good show.

Ok

This is where the nature of federation really shines. .ml is one of the biggest lemmy communities? It doesn’t have to be. In fact as lemmy is still so new and various situations continue to shake out I would wager that in a few years time the top current communities will be replaced. And later on those will be replaced. Censorship resistance is built into ActivityPub by virtue of its decentralization.