I don’t like compose files :)

If you’re proficient it’s 30minutes

Something like this for server.

generate config

docker run -it --rm \
    -v <your-data-path>:/data \
    -e SYNAPSE_SERVER_NAME=<your-public-address-subdomain> \
    -e SYNAPSE_REPORT_STATS=no \
    matrixdotorg/synapse:v1.136.0 generate

run

docker run -d \
  --restart=always \
  --name synapse \
  -e SYNAPSE_REPORT_STATS=no \
  -v <your-data-path>:/data \
  -p 8008:8008 matrixdotorg/synapse:v1.136.0

register user

docker exec -ti synapse register_new_matrix_user http://localhost:8008/ -c /data/homeserver.yaml -u <username> -p <password> --exists-ok

Proxy it using ex. openresty / nginx

location / {
        proxy_pass    http://127.0.0.1:8008/;
        proxy_http_version 1.1;
        proxy_set_header    Upgrade         $http_upgrade;
        proxy_set_header    Connection      "upgrade";
        proxy_set_header    Host            $host;
        proxy_set_header    X-Real-IP       $remote_addr;
        proxy_set_header    X-Forwarded-for $remote_addr;
        proxy_connect_timeout 600;
        proxy_read_timeout 86400;
    }

For UI if you want element on your domain, download and unpack tar.gz from.
https://github.com/element-hq/element-web/releases

Point this location to your proxy server ex. openresty / nginx

location / {
        root /opt/element-v1.11.109;
        index index.html;
    }

Modify config.json inside /opt/element-v1.11.109 to point location to <your-public-address-subdomain>

By default it’s using sqlite if you want postgres or other database then modify homeserver.yaml to use postgres

Thank you for answer. I think I do this one instead https://academictorrents.com/details/30dee5f0406da7a353aff6a8caa2d54fd01f2ca1 Looks like it’s divided by year-month.

How long it takes to download this 3TB torrent ?

You can host ollama and open-webui on container. If you want to wire search you can connect open-webui to playwright (also container) and searxng (also container) and llm will search the web for answers

this is nginx / openresty config - upstream is just definition of server / bunch of servers if you do loadbalancing - you can specify load balancing strategies and stuff. Or when want to separate server layer from proxy layer.

stream {
  upstream something {
     server xxx:123;
     server yyy:321;
  }
  server {
     listen 666;
     proxy_pass something;
  }
}

https://docs.nginx.com/nginx/admin-guide/load-balancer/tcp-udp-load-balancer/

I use openresty with autossl, it renews certificates automatically. The only problem is maintaining subdomain allowance otherwise bots will ddos letsencrypt with random domain names, after some quota they will soft ban you for a week to create certificates for new domains / subdomains.

you can reverse proxy other ports than 443 and ex. upstream ssh, the advantage of having reverse proxy over everything is to have traffic in one place so you can manage it, that’s why for example kubernetes have ingress server, example nginx / openresty upstream ssh, you can restrict traffic to limited amount of IP etc.

stream {
    upstream ssh {
        server          127.0.0.1:22;
    }

    server {
        listen          2222;
        ssl_preread     on;
        proxy_pass      ssh;
    }
}

Buy refurbished / used Mini PC with at least 16GB RAM and 1Gbit network card. Don’t bother with M.2, SSD is enough.
I have right now refurbished hp g4 ryzen 5 mini pc. There is 12 docker containers running there and couple applications in screen, I just add things. Don’t do ARM, as far as I remember running on ARM is compilation hell. Stick with x86. You want it to ease your life and not to make it harder.

Country level internet and passport control before you visit another country domain is inevitable. That’s just like people want it or at least sociopaths.

The future is P2P

maybe some airtable alternative like nocodb or teable

I mean you can try how it works on this website, I believe api is paid for this one but it’s free if you host it yourself.

you can try it here https://libretranslate.com/

searchxng, libretranslate

Have you tried liquidsoap ?

I remember to have it working with default software I got but it was 10 years ago. Then they added this bloatware like media streaming addon or notification center. I have now entware with minidlna and set minidlna scanning disk once per day because media streaming was scanning all the time. I haven’t figure out how to permanently kill their /usr/local/sbin/ncd, ncdb, qNoticeEngined, qulogdb, NotificationCenter, mariadb processes. If I figure out it it probably start working again.

hashed email xD, if someone has email it’s just hash(email) == email. given how many emails leaked producing hashes of 90% of population emails is not a problem

I run this one. https://ollama.com/library/deepseek-r1:32b-qwen-distill-q4_K_M with this frontend https://github.com/open-webui/open-webui on single rtx 3090 hardware 64gb ram. It works quite well for what I wanted it to do. I wanted to connect 2x 3090 cards with slurm to run 70b models but haven’t found time to do it.

I believe you can run 30B models on single used rtx 3090 24GB at least I run 32B deepseek-r1 on it using ollama. Just make sure you have enought ram > 24GB.

If you want to use supercomputer software, setup SLURM scheduler on those machines. There are many tutorials how to do distributed gpu computing with slurm. I have it on my todo list.
https://github.com/SchedMD/slurm
https://slurm.schedmd.com/