The article says they replaced them with remote workers in India, I’m assuming with even lower wages

I’m running Nextcloud from a Turnkey LXC template that’s available in Proxmox. Runs solid, I have no complaints for performance or stability. But upgrades are manual and very involved. It’s not too complicated, but there is always something that needs extra attention or troubleshooting. I also wasn’t able to figure out Turnkey migration toolset that they suggest to use for major upgrades, such as to new version of OS.

@[email protected] has a YouTube channel https://www.youtube.com/channel/UCoryWpk4QVYKFCJul9KBdyw

Well, my original plan was to copy configuration over after I install apos that are not available as flatpaks. Looks like I can copy configuration for those too, just to another location

Interesting. It says that the project is in pre-alpha stage… not sure if I would be able to verify the scripts it generates

That’s what I mostly do now. But it requires some extra work, as some apps are not available in Ubuntu DEB repository. Also, I don’t like the approach that Canonical takes, pushing snaps so much

I really like GNOME. I know not enough about security of it compared to Cinnamon

Thanks! This is helpful

I see your point… I use Debian for my self-hosted environment, so having similar system on desktop may save some cognitive load. My main arguments against Debian are (maybe misinformed though):

• No btrfs support in installer OK, Debian wiki says it’s there
• Major annual upgrades to keep up with stable look more scary than more incremental and frequent updates of Fedora. And using Sid as someone suggested sounds too crazy for main PC

So yeah, looks like it’s just upgrades… Gives me something to think about while I’m moving my apps to flatpaks

Thanks!

Bookmarks and passwords are taken care of. And for the apps I’ll try to get migrated to flatpaks as many as I can while still on original system.

I also see that full disk encryption is being recommended a lot, and I don’t have any solid reasons to encrypt only /home.

I have not given much thought on Silverblue. Is it “flatpak-only”? If so I’ll need to go through my apps to see if that could work. And my backup strategy will need to change - I use Duplicacy that is not available as a Flatpak

Yeah, looks like migration of flatpaks between OS is easy and makes sense a lot

Oh, that’s neat! Thanks!

It’s not just a web front end. I would call it a software development lifecycle service. On top of repos for source code management there could be a bunch of services: Issue tracker, CI/CD automation, static pages hosting, flexible permissions system, even pull requests - all this is not Git.

Forge is a nice and easy name, but not sure if many people realize what it means or recognize that meaning.

Depends on what you want exactly. Easy and self-hosted are not usually go well together unless you’ve got enough experience.

Easiest way for blog - use a platform. WordPress.com is great and has free tier.

More involved, but still relatively easy - static site generator. I use Hugo myself, there is Jakyll that is popular too. Host it for free on GitHub or GitLab pages.

I would not self-host a public web site for security reasons. But you can run a static site on some cloud service. A personal blog with small audience should be fine on Oracle free tier.

Some stuff is in Joplin, some stuff is in wiki.js. Joplin lacks organization features. Wiki.js stores stuff in database and has problems with search, both are possible to fix, I believe…

Occasionally I remember about problems with this setup, but I’m too lazy to fix or replace it

Does this method allow to pick what you need to backup or it’s the entire filesystem?

I can’t say I’m following all release notes, but development is active. You can take a look yourself https://github.com/pulsejet/memories

I used to invent “funny” names, but at some point it became a chore and I also found I’m forgetting some names or spelling when I need it.

Call me boring, but doing enterprise system admin jobs for years I recently started to adopt functional naming convention.

This is what I have now: [location code][OS code][type vm/ct][environment code][workload][index]

So the first production DB linux VM in my primary Los Angeles location will be named LA1LVMPDB1 And my second test Nextcloud container hosted in the same location will be named LA2LCTTNC2.

I still have to invent short names for workload, which is harder for specialized containers, but overall this makes it all more manageable.

Yeah, that’s what I’m using too

No, you can’t