Remember: FLAC files don't lose API endpoints.
The better comparison is PDF + OCR’d text VS Kindle.
But it is lossy, and thus not a good archival format.
At some point, there'll be better lossy formats than OPUS, and then we'll need the lossless source again.
I'd wager that there's a good number of people that don't even own a CD player anymore. $12/month to have music to listen to, even if I don't have new tracks pushed on to me, seems like a reasonable deal to me to not have to deal with the hassle of hosting it on my own server and, what, ripping my CDs to mp3s? If a friend comes in and wants to hear something, Spotify is more likely to have it than my local mp3 collection.
Did you just decide on your 30th birthday that you aren't allowed to listen to new music anymore? I can't even imagine listening to the same music collection on repeat forever.
Me and you are both at least one to two standard deviations above the mean in terms of artist acquisition rates. And Spotify hasn’t been able to help me with music acquisition as much as exploring physical music selections or Discogs, BandCamp, or eBay has, and in those cases I end up with higher quality listening experience as well.
e.g. mine are in a well backed up filesystem, reachable from anywhere in the world via my tailscale network.
Typically mounted read-only on whatever computer I am using, at definitely less than $12/mo.
In contrast, Spotify doesn't even have much of the music I listen to.
My bicycle won't run out of gas but that doesn't mean it's more useful/reliable than my car for a cross country road trip
That way, you can have the information you need w/o being tracked.
Do you have a solution for how to cheaply extract features from songs? If yes, I'd love to hear about it, but if not, your evangelism and impotent attempts to defend it are not productive to this conversation.