vegetaaaaaaa

This is the correct solution for unattended file copy, plug the external drive and wait till the copy finishes. The only difficulty is monitoring progress/completion/failure if that’s on a headless machine. I built a basic system with systemd .path/.mount/.timer/.service units, logging and monitoring around this, once you get it working, it works.

The free edition of Mattermost is artificially crippled [1]. Amongst other things, only the most recent 10k messages will be displayed, older messages will be hidden (borderline ransomware). They took a bad turn when they decided to start LARPing as a US defense contractor company (I mean look at this homepage [2], wtf). I wouldn’t recommend it for a new deployment.

For document scanning I use this f-droid.org/en/packages/org.fairscan.app/ directly on the phone, works well even with a shitty camera.

I never had the need to print from android but I found this f-droid.org/…/io.github.benoitduffez.cupsprint/, it might work.

Scanservejs looks nice but one more service to maintain and secure. I guess I’m a minimalist.

I use SANE for scanning.

• install sane-utils
• configure /etc/sane.d/saned.conf, list the IP adresses or network allowed to connect
• start saned.socket

on your client (linux desktop)

• add the SANE server IP to to the config echo “192.168.xx.xx” > /etc/sane.d/net.conf
• list the scanners scanimage -L

The scanner is also automatically listed in scanning utilities like simple-scan etc. Don’t know about phones or other devices, I assume there are scanner auto-discovery utilities. I don’t know about Web UIs, just use native printing/scanning features of your OS.

FUD

Gitea is still MIT licensed and the Enterprise tier features only cater to large org needs [1]

why would I want to deal with nightmarish SAML config when OIDC does the job

• Small 4B models like gemma3 will run on anything (I have it running on a 2020 laptop with integrated graphics). Don’t expect superintelligence, but it works for basic classification tasks, writing/reviewing/fixing small scripts and basic chat
• I use github.com/ggml-org/llama.cpp in server mode pointing to a directory of GGUF model files downloaded from huggingface. The use it from the built-in web interface or API (wrote a small assistant script)
• To load larger models you need more RAM (preferably fast VRAM/GPU but DDR5 on the motherboard will work - it will be noticeably slower). My gaming rig with 16GB AMD 9070 runs 20-30B models at decent speeds. You can grab quantized (lower precision, lower output quality) versions of those larger models if the full-size/unquantized models don’t fit. Check out whatmodelscanirun.com
• For image generation I found github.com/vladmandic/sdnext which works extremely well and fast wth Z-Image Turbo, FLUX.1-schnell, Stable Diffusion XL and a few other models

As for the prices… well the rig I bought for ~1500€ in september is now up to ~2200€ (once-in-a-decade investment). It’s not a beast but it works, the primary use case was general computing and gaming, I’m glad it works for local AI, but costs for a dedicated, performant AI rig are ridiculously high right now. It’s not economically competitive yet against commercial LLM services for complex tasks, but that’s not the point. Check old.reddit.com/r/LocalLLaMA/ (yeah reddit I know). 10k€ of hardware to run ~200-300B models, not counting electricity bills