What does Jellyfin have to do with that? If you implement acess control in the reverse proxy, requests from non-whitelisted IPs are just not forwarded to Jellyfin.

I have mine behind a revwrse proxy (Nginx Proxy Manager), and use a whitelist to allow specific IPs or IP ranges access so my family can use it.

30-40 billion USD in total worldwide over three years seems very little compared to the massive expenditures by the AI companies to build the things?

They rely on it, that does not necessarily mean that they have written it. I have never heard of Murena being involved in the development of Nextcloud before.

How do you like it? We’ve set it up as a test at work, but we don’t really do much office-type document collaboration, so we haven’t really tested it much yet.

Murena is also the author of NextCloud

What? I don’t think this is accurate?

I use this for archiving news and magazine articles as well (with snapshots), sorted on topic so that I 1) might be able to remember where I read something and easily find an article again if I discuss it with someone and 2) have a good starting point for researching something I don’t have time for or the will for now.

I have set up the file sync on a self-hosted WebDAV server as well as it quickly racks up storage space with all those snapshots and you fairly quickly reach the top tier storage plan they offer.

Zotero 7 brought some good UI improvements, but it is really resource heavy (at least on Linux). A CLI-interface as was mentioned under here would be interesting.