IronKrill
@IronKrill@lemmy.ca
- Comment on Selfhosting Sunday! What's up? 2 weeks ago:
Giving away? As in free? Luckyyy
- Comment on Selfhosting Sunday! What's up? 2 weeks ago:
The best thing I did was probably set up Obsidian Live Sync after hearing about it in selfh.st weekly. Compared to the Syncthing setup I had before it is so much quicker, much easier to catch sync errors, and won’t drain my damn phone battery as much! The more things I can move away from Syncthing the better so I will probably be looking at Immich for my photos soon as well.
On the management side, I’ve had my Docker stacks split across Dockhand and Komodo ever since a poorly executed move away from Dockge a few months ago. I finally thought about it and had the time to decide to give up on them both. I like some of the concepts of compose in a git repo and the fancier interfaces, but I just realised it was all over the top for what I need. I missed being able to simply SSH into a damn machine and
docker compose up -din an emergency. Compared to the constant deploy errors of Dockhand and the potential of scrounging application databases for in-DB compose data or attempting to deploy stacks from forgejo when forgejo itself is a stack deployed via git… going back to basic files is a blessing. Lesson learned!What that culminated in is all my stacks are now in Sencho (which clearly uses a lot of AI code sadly), but it was that or Dockge and I knew Dockge lacked a few things I wanted. The nice thing about Sencho (and Dockge) is that I don’t need either of them because they are based around file-on-disk simple operation. I can cut it out of my setup entirely at any moment and not shed a tear. So much peace of mind leaving Dockhand. Now I just need to find a builder image or write a script to replace Komodo’s build feature because I don’t want to keep using it.
- Comment on Jellyfin 10.11 RC1 Released 1 year ago:
Media watch status no longer persists on media rename. When you move or rename your media files, previously Jellyfin remembered the watch status based on external IDs alone; now it uses an actual reference to the database entry. This is a transitional change and we are working on a better way to handle that in a future version.
I hope they do figure out a solution for that soon… Until that’s fixed I’m going to stay on 10.10.7 for a while.