i have red and still got told to mute while writing

the power of an annoyed programmer

if you configure homepage with docker labels, and have multiple homepage instances, they will all be the same. since there is no way to specify the instance on the label.

reference:

• https://github.com/benphelps/homepage/pull/776
• https://github.com/benphelps/homepage/issues/774
• https://github.com/benphelps/homepage/discussions/1261

not yet, it is planned but not there yet

you got me laughing here

thank you for sharing.

and i’m definitely spoiled too, my previous job paid me 35k gross per year and i want no less than that. i agree with you that software engineers here are under paid, but it is deeper than that. the software engineer job (and derivatives) does not exist in Italy. the contracts are generally or extremely generic (and so the pay brackets are reeeally low) or you have a contract for something totally unrelated (a friend of mine work in IT and have a contract for mechanics).

i got spoiled because my first job was from a German company, and their pay is a bit more fair than the ones in Italy.

funny how here in Italy 30k gross per year is a pretty high salary.

software engineers with required experience in a field can get paid as low as 20k and they also want you to actually go to their office.

is there an advantage to have apps embedded as iframe? as opposed to opening the url in a new tab?

like forwarding auth?

the only annoying thing is that is not possible to spin more than one homepages at the same time.

so i have one homarr and one homepage

well i don’t think you are supposed to. is still full beta, no debugger.

i think this is just for their community to test stuff on.

they can detect piracy regardless. generally they set up honeypots and if you download a torrent from them, they will know and fine you.

i do it… and that’s bad right?

for docker the syntax is --gpus all

https://docs.docker.com/config/containers/resource_constraints/#expose-gpus-for-use

bonus: syntax to expose the gpu in a docker compose

    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]

more at https://docs.docker.com/compose/gpu-support/#example-of-a-compose-file-for-running-a-service-with-access-to-1-gpu-device

i have a public ip and my own domain attached to that. i use subdomains for each service and a dashboard on the root domain.

i don’t use authelia etc, and rely on the autentication page of each service. but i have fail2ban.

i did help them set up the apps, but they took from there. the dashboard on the root domain helps them navigate all services without having to remember the full url.

yes… maybe.

as the dev said, it flags a lot of false positive. so a human should look at them anyway.

maybe when this is a bit more evolved, we can use it to preprocess posts, and if a post gets flagged for something, a mod / admin needs to approve the post manually.

maybe for CASM, it gets sent to an external service specialized to that stuff, so the mod / admin doesn’t have to look at the images.

it uses a model that describes a photo, then it searches the generated description for some terms and ranks the image to some levels of safety.

to test it you use a more general filter, for all nsfw for example, and see if the matches are correct.

if it helps, here is my setup from bare metal to 30+ services. https://github.com/simone-viozzi/my-server

including off-site encrypted backups

of course, i wrote this for me, so most stuffs are written like garbage, but fell free to open an issue, and i will fix them

this is the way.

to run on ROCm, you need a specific version of pytorch.

but it is still in beta, i would not expect it to run well

the model under the hood is clip interrogator, and it looks like it is just the torch model.

it will run on cpu, but we can do better, an onnx version of the model will run a lot better on cpu.

check the transcoding section on the server and on the phone

Available from internet:

• jellyfin
• jellyseerr
• immich
• paperless-ngx
• owncloud ocis
• traefik
• homarr

Available only from local:

• the *arr stack
• qbittorrent
• jackett
• watchtower
• apprise
• netdata (kinda new, still have to fully understand how it works)
• portainer
• speedtest-tracker
• homepage

Security

All the services available from internet, just goes through traefik to terminate https, I rely on the build in authentication of each service. To add another layer of security, I have fail2ban active on all those services.

I have a public IP, and I have open on my router ports 80, 443, a random port for ssh and vpn.

Hardware:

Memory:
  System RAM: total: 8 GiB available: 7.73 GiB used: 4.46 GiB (57.7%)
  Report: arrays: 1 slots: 4 modules: 2 type: DDR3
CPU:
  Info: 6-core model: AMD Phenom II X6 1090T bits: 64 type: MCP cache: L2: 3 MiB
Graphics:
  Device-1: NVIDIA GP107 [GeForce GTX 1050 Ti] driver: nvidia v: 535.98

docker compose files

All the docker compose files + how I configured everything is available at: https://github.com/simone-viozzi/my-server

Bonus:

Since I like the ability of btrfs to do snapshots, I created all important docker volumes as btrfs subvolumes. Then I created a backup script that literally sends the subvolume (encrypted) to an external cloud. This does not allow incremental backups and most likely is not the best backup solution… but it works… the repo is: https://github.com/simone-viozzi/btrfs2cloud-backup

I welcome any advice / criticism!