And I consider this more successful, if it is based on existing work and just adds the federation there. Because I think it is a lot of work to do correctly and that has already been done. And all that work is not related to the federation aspect.
My wild guess would be scaling issues stemming from principal assumptions like "there will never be more than 255 people on the same map" and "all access will be unauthenticated and ad-hoc anyway, so why bother validating user input".
Also, seeing basically only NixOS content in the repo, I gather that the "from scratch" idea rather refers to "we develop with deployment in mind, so let's start with the deployment environment, then integrate, re-create or create what we need"
@rugk Better ask the maintainers why things are done as they are done. However, by now it seems that (parts of) the WA source code might also be used, as they pull in a repo called like this: https://github.com/SuperSandro2000/workadventure-nix
Personally, I like the idea to use Nix as the deployment environment. For everything else, it's to early for me to say whether the choices are good.
@rugk @anathem FYI: Some more planning on how the federation should work: https://gitlab.com/fediventure/fediventure/-/blob/main/pages/fediventure-net/content/specifications/contents.lr
chaos.social – a Fediverse instance for & by the Chaos community