20 Nov
2021
20 Nov
'21
1:22 p.m.
Unfortunately, we're just not doing enough. I
don't have an answer, but for several years I worked for the LOCKSS project at
Stanford University. Their software is open source, and is used to archive online academic
journals in a distributed, fault-tolerant peer to peer network (their primary research was
on how to prevent bitrot and enable automatic format migration, for example). With some
effort, it could be retooled into a general-purpose software archive, but I don't
think anybody has ever seriously proposed doing that. I only point it out as inspiration
for one possible way to do things.
Until we have a better answer, we're just at the mercy of bitrot, and we have to
accept that.
Unless, of course, you have a mainframe shop and figured all this shit
out back in the 1960s.
--
Will