You could probably do this with FUSE. Guess nobody cared to make that yet.

Opencloud is a fork of the new Owncloud, I think. Similar to how Nextcloud was forked from the old Owncloud.

You can access all Nextcloud files over WebDAV. That is natively supported by many file browsers, including explorer.exe on Windows.

And you can choose in the Linux client what folders to sync.

What the Linux client (in contrast to the Windows client) does not support is having virtual files in a folder and only downloading files on demand.

Apart from that, have you looked at Opencloud?

Other way around. Sunshine is the server, Moonlight is the client.

I so so love open source. “We are all in competition with each other. Let’s pool our resources and share knowledge to make us all better.”

10.0 is tracking the Wine release it is based on and -4 is their own iteration on that version.

That’s perfectly fine. You can mix and match. Only downside is that you get all the dependencies from different DEs as well.

Nice!

Sounds like something that would be well suited for integration into umu.

Not sure, but does removing the meta package and then running apt autoremove work?

Gwenview when it’s really small.

I had it running on my Vega 64. But it had to be exactly one specific version of ROCm. Been a while since I’ve played around with that so I don’t remember the specifics.

With the way they were talking I thought it already WAS working and that this PR just published the code.

I switched to rspamd. Its bayesian filter is a little weird. It only started working ok after I found the right amount of mails to feed to it. For some reason it forgot everything if I gave it too many mails. I think it’s a Redis thing. No idea. I don’t have the brain power to figure it out or write a proper bug report. But I think my Debian version is outdated anyways, so this might be fixed by now.

For my server learning from mails from the last 50 days was the sweet spot. Since then I got no false positives and only the occasional false negative. Exactly how I want my spam filter to be.

The whole drive. The docker file and volumes are the bare minimum.

In general you backup everything that cannot be recreated through external services. So that would be the configuration files and all volumes you added. Maybe logfiles as well.

If databases are involved they usually offer some method of dumping all data to some kind of text file. Usually relying on their binary data is not recommended.

Borg is a great tool to manage backups. It only backs up changed data and you can instruct it to only keep weekly, monthly, yearly data, so you can go back later.

Of course, just flat out backing up everything is good to be able to quickly get back to a working system without any thought. And it guarantees that you don’t forget anything.

Yes, let’s keep taking features away. 🙄

This is actually an accessibility issue. It’s often much easier for me to use middle click paste than other copy and paste methods. But as always those numbnuts just think about streamlining everything.

The trouble with pictrs is that it sorts pictures into seemingly random folders.

The solution is to not proxy images. Might even be the default by now. That’s a huge resource hog. No idea what pictrs is doing but it’s still taking up a whole lotta space just for my own images.

Canceling all subscriptions would probably make Lemmy use almost no resources.

I run a single user instance and it’s horribly slow. Mostly because I only have HDDs and not enough RAM to compensate. I hope Lemmy 1.0 will increase database performance.

Piefed is supposedly much more performant. But I’m shying away from migrating because I don’t want to lose my post history and uploaded pictures.