I’ve always been conservative about what kind of services I host because it takes time to get them set up. For example, there’s no reason for me to set up music streaming when I only ever listen to music on my phone and all my music files are already on my phone. On the other hand, it’s a good learning opportunity to set stuff up and have to fix it when it breaks. What do you think?
I spent a lot of time setting up firefly-iii, a really neat and feature-rich finance manager. It’s a really great piece of software by a very responsive and friendly dev but after about 6 weeks I still couldn’t get used to it and ended up going back to paying for YNAB.
I swear by memos now though - highly recommended. It’s like having a private twitter stream where you can send thoughts, notes and files that you want to store/refer back to.
I spun up Firefly a few months ago and had about three weeks where I was actively categorizing transactions and reconciling everything and then my ADD kicked in. Really cool tool but I just need something low-maintenance for budget tracking.
I really liked Firefly III but it doesn’t allow negative budgets, so I’m running Actual Budget now.
How do you like Actual? I set up Fireflyiii as well, but once I read that there is no way to share a ledger, so to speak, it turned me off a bit.
My wife has bookkeeping experience, so something that is a bit closer to double entry bookkeeping would be awesome, since it should fit easily into her quickbooks experience.
Currently looking at akaunting, which seems like it may work, if it is truly self hosted
I’ve been able to make it work for my wife and me. We don’t distribute income as it comes in, so I’m ignoring like half the numbers in the UI, but it’s working.
I’m in the US, so there’s no good self-hosted way to get access to my own financial data, so I’ve got all my credit cards and my bank account emailing me alerts, and then I’m parsing the alerts into Actual. I’ve also got budgets filling automatically using schedules.
Once upon a time I set Grocy for my wife and I. It’s a kitchen inventory management site that can do some pretty cool things. I spent a whole weekend scanning every barcode in our kitchen and getting a bunch of recipes setup on the site. After all that we used it for maybe two or three weeks before it felt like to much work lol.
Do you remember if you were able to pull up items based on barcode? I’ve seen that you can record items based on barcode, but what I mean is: if I’m eating ice cream, can I scan the barcode so I can update the quantity quickly? Been searching their docs but can’t find anything 😅
If something interests you, set it up. If you find you don’t need it, take it down.
That’s pretty much exactly what I have done. I’ve hosted Plex, and Matrix in the past. Plex I will host in the future but Matrix was too much for me to host on my own, but the experience of setting it up myself was definitely worth it.
I recommend checking out Jellyfin instead of Plex. Open source, fully self hosted.
Can I ask why the Matrix was too much? I’m thinking of setting up a Synapse docker container.
Well I don’t have a home server at the moment so I was renting out a server for about $40/month. Mostly because it was eating up a bunch of storage, and I was too lazy to swap from digital ocean.
Also I was using the ansible script and at some point I they changed something that required me to set it up again which I didn’t really have the time for.
I will say Ansible was a lifesaver. It made setting up and keeping the server up to date super easy.
I do recommend trying it out tho, just don’t use a domain name that is the same as your username or you will have issues with pings, especially if you share the instance with a friend. Learned that the hard way. Anytime they sent a message anywhere I was at, it pinged me, whether or not they intended to ping me.
Oh I have quite a few that I’ve set up then pulled out of service for various reasons. I’m always evaluating potential use-cases for new services and if a different service would better suit my needs than what I have deployed currently. It’s definitely a hobby.
Some container-based projects that I’m loosely tracking updates for and have deployed, but since, have pulled out of service (non-exhaustive):
Media:
-
- Supplanted by Audiobookshelf. For tagging and book conversion I just temporarily install Calibre on my Workstation when the need arises.
-
- Did not end up using it for much at the time. Re-evaluating if I’d like to stand it up again.
-
- Plex is nice but too many drawbacks that don’t work for me. Supplanted with Jellyfin.
-
- Didn’t have much use for this, but it will likely change soon. Since I’ve stuck with Jellyfin I’ll be going with Jellyseer if I decide to stand up this kind of service again.
-
- Rarely made use of it. Nice project, but it’s not feasible now for obvious reasons.
-
- Same as FreshRSS, though I’m a big fan of go and rust projects in general so this is the one I’m more keen on re-implementing.
-
- Part of the Plex ecosystem which I abandoned. It was useful software, but unfortunately locked to Plex.
-
- Never really had a use for this, though I thought I would at the time.
-
- Project is decent but the author is a asshole and very user-hostile, so I dropped it when I retooled my homelab a few years ago.
-
- Supplanted with Prowlarr.
-
- Same as the reason for dropping Overseerr.
-
- Did it’s job, consumed lots of resources, and no arm64 docker image; though I managed to build my own. Got rid of it when I no longer had a use for the service.
Archival/Documentation:
-
- Something about the project rubbed me the wrong way, vague on the details though.
-
- Was decent but the lack of updates then subsequent maintainer turnover scared me off. I check in from time to time to see how the project’s going.
-
- Ended up being too slow and clunky for me, but that could be the hardware I was running it on at the time.
-
- Same as above, but it definitely wasn’t the hardware.
-
- It was alright but decided I didn’t need a separate service for documentation, I just use Code-Server with a documentation repo and raneto to give me a pretty page to navigate and for the family.
-
- Worked well while I was using it, not a fan of the closed-source nature and just didn’t feel the need to redeploy when I retooled my infrastructure.
-
- Same as bookstack, didn’t really have a use for a separate service.
Dashboards:
I’m going to preface this by saying I have some sort of addiction with dashboards, it’s unhealthy really.
-
- Didn’t like how everything was an iframe and it seemed particularly resource heavy for what I needed it to do.
-
- My second dashboard. Liked the API integration not so much the design.
-
- Wasn’t a fan of the design.
-
- Also didn’t like the design much.
-
- Decent project, but decided to move on to something configuration file based.
-
- Liked this one a lot and used for quite a while before homepage lured me away with API widgets.
Infrastructure:
-
- Inadvertently made my infrastructure brittle with how I had it implemented. Decided to just rebuild my cluster’s cloud image on-demand instead of daily and update my apt distros the old-fashioned way.
-
- Liked this project a lot, but added complexity to my infrastructure that could be more simply achieved other ways.
-
- Same as lldap, but more feature rich and thus even more complicated.
-
- Set-up briefly but found a better use-case with Gitea’s integrated package registry which I’d already had deployed.
-
- Used this for a while, but the clipboard situation sucked at the time and I gravitated to just using SSH anyway, and since I have Proxmox on hypervisor duties just used xterm.js or noVNC for console access.
-
- Did it’s job but the
:latest
tag is dangerous to use. I like having change logs, an evaluation environment, and an approval based update workflow so I switched to renovate-bot.
- Did it’s job but the
-
- Was a decent option for sure and faster than what I’m using currently in theory, but seemed a little to unnecessarily complicated to keep running for me.
-
- Definitely useful. Will probably redeploy it at some point.
-
- Supplanted with Wireguard implementation in Tailscale.
-
- Tried it very briefly but N8N fit my use-case better.
-
linkding was one for me. It sounded like a great idea at first, but, I never used it. Shut it down after a couple months.
I would rather waste a week setting something up to find I don’t like it, then paying some company to give me some ad riddled thing that phones home every few minutes and being stuck with it for a month, then the nonstop emails after I’ve cancelled and my information being sold to who knows who.