SearXNG doesn't work

[papu163]

I used the python method for linux to get Odysseus but when I go to localhost:8080 SearXNG does not appear and because of that SearXNG doesn't work on odysseus

Nevermind just started searXNG through docker Im so stupid

SearXNG just refuses to work with the AI, I set the default engine to SearXNG and the fallback to duckduckgo and the AI uses for some reason duckduckgo instead of searxng(searxng works in my browser)

[pdlzz]

How do you see that SearXNG is working? Im using the native windows route and when i run the test in settings it shows "active - 5 results" is that correct?

[papu163]

Im on linux but Im sure it works on windows too, just go to http://localhost:8080 if there is no connection then it doesn't work but if SearXNG appears then it works

And yeah the test button is correct but It doesn't work for me most of the time so I just connect to http://localhost:8080 to see if it works

[pdlzz]

I managed to get SearXNG working by giving it a key in the settings.yml file. Websearchs work as intended. No port changes or anything.

[Ant-V]

This looks to also be a problem on macOS (https://github.com/pewdiepie-archdaemon/odysseus/issues/2600)

[maddes8cht]

My main problem with SearXNG is that it is running on port 8080.
So i changeed it to 8060 in the 'docker-compose.yml' file and hope that this is all it needs.

I'm using Odysseus with my own llama.cpp installation that is used also by a bunch of different other clients that all expect it to run on port 8080.
8080 is the default port of llama-server.
I know you can install and run llama-server from inside Odysseus and i guess it will use another port there which is easy to do, but: For me Odysseus isn't the only client that is using llama, so i prefer to run it myself independent from odysseus, and i guess there might be a bunch of other people around doing the same.

This Project is still in an early stage.
I would sugest to change the port for SearXNG to not get future unecaessry conflicts for people using their own llama-server installation.