This is the version I setup just yesterday. Much simpler setup than the AIO. The AIO controls Docker to manage its collection of containers.

Shitty ass thin client running cheap hw that can’t do anything, a.k.a. Chromebook.

And when they do, you take care of it. Also, if you use Tailscale or equiv, you can stop Syncthing’s exposure to the internet. Then you can stay on a fixed version across clients and limit unexpected breakage that comes with autotomatic updates.

Build a Syncthing Android apk yourself. You don’t need to update to every release. I’m still using 1.30 with 2.x.

Is there difference in how much storage space is needed between the two approaches?

How does this compare to redarc? It seems to be similar.

The process through which we get more users is that something material changes in current Tirefire user’s life that puts them over the threshold needed to look for alternative. Then they look. Lemmy is the obvious Reddit alternative, it’s well indexed in search engines. Then they try it. If the quality of content is decent, there’s a decent chance they stay. They know the quantity won’t be as high, that’s the major reason they haven’t switched to begin with. So for this process to keep functioning, we need to maintain the quality.

Of course we should also suggest Lemmy, but probably when asked or otherwise appropriate. Or else it may have the opposite effect that naked shilling often has.

This is going to be a bigger deal than I thought. Can’t cross 91% lib scan even if I delete the music lib and then upgrade. Even turning on debug logging doesn’t reveal anything helpful. Rolling back to 10.10 for now. Will attempt again when they force me to upgrade.

Good note.

I seem to be suffering from this issue. Lotsa spam in the logs and the scan job doesn’t seem to be moving. The next attempt would be, revert, disable music lib and upgrade. Then deal with the music lib separately. But I’ll let it run another 12-24 hours.

My god what a shit show.

It’s all normal though. Software development is hard and big changes mean more regressions. You do the changes when needed, then you work on squashing the bugs.

Library scan still going here. At least something is spinning because one CPU thread is pegged to 100%.

I do. It could be this:

After the migration it is recommended that you perform a full scan through the admin dashboard. We have observed that for some users, some elements might not work properly otherwise (e.g parental ratings). As of RC8 a scan for missing metadata may be required for music libraries to function properly. The first scan after the migration might also take quite a bit longer than usual, though subsequent scans should be as quick as before.

From the release notes.

The last messages in the log are from music lib scanning. I’m leaving it be for now.

Fucking hell. OK, I’ll upgrade again and wait longer. Mine’s ~15TB.

I do have a large collection and I haven’t checked the music lib state yet. I have a zfs snapshot from before the upgrade so I could restore if shit hits the fan.

Just upgraded from 10.10. Seems to have gone well. Library scan still going.

E: Couldn’t finish lib scan. Reverted to 10.10.

Interesting.

Why terrible?

So far no one who’s able and willing to do the work for a fork has done it so I assume the motivation is low. There’s alternate pieces of software mentioned here that have their pros and cons.

I donate monthly (recurring) to:

• Lemmy development: https://join-lemmy.org/donate
• Our Lemmy instance: https://fedecan.ca/en/donate

Currently most large instances are well funded so I’d personally focus on development funding.

Been using Jenkins since before it was called Jenkins. It’s been in use at every corpo I’ve worked for. It can practically do anything. Especially coupled with Docker.

Jenkins