I used umami cloud (free tier) on my personal site for 6-12 months and can recommend it. I ultimately decided to switch to just a simple counter for visits and likes because that's all i care about and i don't like collecting more information than i "need". Now my website has no other tracking/analytics and the entire site still works even if Javascript is disabled.
For a business if you're wanting a Google analytics alternative then umami does a great job. Self hosted option available as well.
To your initial question, if you'd actually use the data then i would recommend some form of privacy respecting analytics on your business site though. Since you seem privacy conscious i just wouldn't put more analytics on your website than you'd personally be ok with as an end user
(I can't see the edited out part but if it was about yay...)
Yay builds in your local cache and then when it is ready to install it asks for sudo. The reason for this is because sudo can timeout during long builds, and more importantly if you compile with sudo you run the risk of arbitrary code execution. So it is safer to run with just yay and then it will ask for sudo when it actually needed.
Hugo (markdown) files that i host on my internal server.
how you organize it
I use basic directory structure. Top level directories are like "dev", "home", "general". Self hosting is a dev/ subdir.
what information you include
Depends on how familiar i am with it and how often I'll be referencing it. Something i know well or access often will be more high level. Things like an annual process i have documented in more detail
how you work documentation into your changes/tinkering flow
My site has an "edit this page" feature which i use to open my IDE and make the change as I'm doing things. Sometimes I'll be lazy and just add in what i did this time and then let future me reconcile the differences 🙃
To be fair MS makes orders of magnitude more money and has the benefit of operations at scale. Whereas codeberg's operational budget for 2025 was 100k euro and they still need to deal with DDoS and bot scraping. They also were running off a single server up until sept'25 when they had two donated hardware services which are now hooked up to make a 3 node ceph cluster.
If I'm being completely honest, it sounds like you hit a problem and then just kinda gave up (I'm not trying to sound mean or anything - please don't take it that way).
I got the same impression. Which is fine if that's someone's approach, but that same person probably shouldn't be on an arch-based distro if that's the case.
Yup! Mostly symfonium since i mostly use my phone for music. Started using feishin recently for desktop use and have been really impressed with it. I can recommend both!
for music both jellyfin and navidrome are subsonic API compatible for use with mobile and desktop apps (like symfonium and feishin). Some people choose to just use jellyfin instead of a dedicated music service. Personally i still run navidrome for music. I give some thoughts on that here
This is what i do. Have certbot running every night, and it'll auto skip if it is too soon to renew. If renew is successful then it'll deploy. Pretty much set and forget it.
I also dropped strava a while ago. For me it was because they updated their privacy policy to blanket allow ai training with your data to both strava and any partners. They claimed it was only for XYZ but the privacy policy allowed it for any use which i consider dangerous for health and geospatial related data without specific, informed consent.
But for alternatives, when i was into cycling/triathlons i used golden cheetah extensively. It's UI takes some getting used but ime it was more powerful than anything else once you got used to it. I used it as a strava premium/trainingpeaks premium alternative and had multiple athletes (me+coaching) in there.
I feel like you didn't read the post or issue i linked, nor their license.txt and are instead just trying to talk past me.
I don't really care about this project or debating their intentionally ambiguous license structure. My point was that the grant of rights explicitly only grants AGPL access to create compiled versions of mattermost. That is not how FOSS licenses work and is incompatible with FOSS licenses because it lacks the "freedom" that even AGPL would typically grant.
You may be licensed to use source code to create compiled versions not produced by Mattermost, Inc. in one of two ways:
Under the Free Software Foundation’s GNU AGPL v3.0, subject to the exceptions outlined in this policy; or
I'm not saying that people can't dual license or that they can't release their product in other non-free ways. That's not the issue here. The issue is that you are saying it's AGPL, and it's not--Not really. It's only AGPL to create a compiled version of mattermost.
Might be worth reading this and the original github issue. It isn't actually agpl. They only grant access to the source code to build a compiled version which isn't freedom. And beyond that, some code is covered under a source available enterprise license which i think is where they would enforce their paywall
Some form of hosted musicstreaming integration with my local music
For the music, jellyfin can do this and it uses subsonic api which means you can connect to the music server with some mobile and desktop apps. Alternatively i like navidrome for more specialized music service that still uses subsonic api. Some people prefer not having a second service if jellyfin is good enough for their needs.
Automate Backups and push them on my server
For backups look into borg if your NAS doesn't have anything native.
make all of the above things available where ever I want using my own self hosted domain.
Look into doing let's encrypt DNS-01challenges via something like acme.sh if your domain registrar has an api. this will let you get your own certs for local use without exposing the subdomains on the domains dns. If you're going to make them public then that is less important but it's still a good way to automate renewals and deploying regardless.
run my own dns
Pihole unbound can offer a recursive dns server. Very easy set up.
In the long term I also want to be able to host my own webapps, since I will soon start to develop one for someone.
Now I want to know what suggestions do you have, for stuff thats really cool and that I can selfhost.
Outside of the obvious segmenting public zones and firewall, you could self host an SSO service. This would allow you to easily put forward auth on a dev build if you were needing to keep it selectively private until/if you made it public.
In general though, i just wait until i come across a problem or need and then i see if a service exists to solve that. Occasionally looking through the awesome selfhosted list or similar helps find blind spots i didn't know i had.
Some questions to help explore if i was in your shoes:
Are these internal playlists (db) or external playlists (m3u)?
if external playlists- do they list the full filepath or just the relative file path? (Youd want relative)
does the user have access to both libraries? from the docs:
Playlists: Can contain songs from multiple libraries (user must have access)
Smart Playlists: Can be scoped to specific libraries using filters
I haven't experimented with multi library support but that first bullet point makes me think that (db) playlists might be tied to the original library song, and navidrome is expecting the user to have access to that library to play those songs even if the songs exist separately in the current library via symlinks. The second copy of those songs are likely treated as second entries in the db even if they ultimately are the same file.
An alternative approach would be to create a backup of the db, and then delete the symlinks, delete the "missing songs" from the navidrome ui, and then move the song files to the new library. Right now you're thinking of doing a parent library (all files) plus a child library (selective files from parent), but id treat the default library as "core" songs everyone should have access to, and then a second "personal" library for your specific songs. The caveat is i don't know if navidrome will be able to tell the songs moved from one library to another. If not, then this could be pretty messy if you're wanting to keep annotations and playlists which is why i suggest a backup.
I used umami cloud (free tier) on my personal site for 6-12 months and can recommend it. I ultimately decided to switch to just a simple counter for visits and likes because that's all i care about and i don't like collecting more information than i "need". Now my website has no other tracking/analytics and the entire site still works even if Javascript is disabled.
For a business if you're wanting a Google analytics alternative then umami does a great job. Self hosted option available as well.
To your initial question, if you'd actually use the data then i would recommend some form of privacy respecting analytics on your business site though. Since you seem privacy conscious i just wouldn't put more analytics on your website than you'd personally be ok with as an end user