3 ms·
ipfs might be a hosting solution, but that's a small part of the challenge. You also have to source the data. I recently used some online periodical archive ac
by lkbm 4y ago
ipfs might be a hosting solution, but that's a small part of the challenge. You also have to source the data.
I recently used some online periodical archive access from my local library to find sources about 1940s musicians protesting recorded music...couldn't find anything. Then I tried IA, and boom, they've got it[0] — someone had to source and scan all these.
I recently wanted to revisit some 1980s-era computer games from when I was a kid. Internet Archive. How'd they get there? Someone there pulled them from 5.25" floppies.
A few weeks ago, I remembered 3-2-1 Contact magazine, which I enjoyed as a kid. No problem: IA has source and scanned them.
I could store all the DOS games in the world on my hard drive without noticing the space usage at all, but step one is acquiring copies of them. When it comes to the long tail of obscure 1940s periodicals or 1980s advertisements[0].
[0] https://twitter.com/lkbm/status/1608480435691995137 https://twitter.com/lkbm/status/1608480435691995137
[1] https://twitter.com/textfiles/status/1643781181568327681 https://twitter.com/textfiles/status/1643781181568327681
- hirako2000 4y agoI agree the sourcing isn't something ipfs is magically solving. It is specifically to reduce the weakness of more centralised data store/serving which are easier targets. Even in term of resilience ipfs isn't bullet proof. Decentralising helps. Pinning needs to exist somewhere as a hard rive and a network connection, ideally with multiple replication, and held with serving as such for extended periods, across every single file.