This thread by Zooko is worth reading to consider just how much we depend on a *single* institution to keep most contemporary cultural production alive. Yet much is *not* archived (eg YouTube!) https://twitter.com/zooko/status/1271151348273143808
The fediverse has not yet done better, but could (I left a rant here:) https://twitter.com/dustyweb/status/1350938649454669830
In general, content-addressed storage seems to be most suitable for long-term preservation of data because file-names and URLs do change all the time, and the biggest modern tragedy is when you find some image on on some ancient forum, but it's no longer there - you have its link and filename, but that rarely helps. Yet, you can be certain the actual image sits *somewhere* in someone's cache, you just can't discover it...
