I have one for categorizing subscriptions/channels. I've been running it for 3y maybe more. SQLite too. No history integration. I have not opened sourced it because the code is thrown together and there are some subscriptions that the YouTube API doesn't return. I'm not certain what the commonality between them is, either. Possibly country of origin.
My impression is that costs will continue to go down. Large investments are unlikely to be profitable for these businesses. Whoever is dumping billions into this is unlikely to get their money back. The new tooling, models, discoveries seem to be commoditized within months. There are no moats. If things keep going this way there will never be a point where employers (or anybody for that matter) have to pay the real cost.
The simplest setup is to run jellyfin and tailscale in docker with docker compose utilizing tailscale serve. You will get automatic https and a simple reverse proxy with tailscale serve.
Programmers don't fully control what they work on when employed by an entity seeking profits.
Programmers _do_ have control in the world of open source. Unfortunately efforts are spread out thinly[0] enough to prevent many ideas from reaching the tipping point to being better than a profit driven entity's solution.
Imagine what would be possible in OSS if all work in a similar domain was concentrated.
I have seen this happening at work. What should have been a single file one-off script to do something was instead a full blown application "just in case".
I agree with you on the sleep-hibernate thing. Unfortunate that Linux sleep is noticeably worse. Switched from Mac to AMD Framework 16 with Fedora and I discovered it quickly.