Conversation

Notices

1. Embed this notice
   Nemo_bis 🌈 (nemobis@mamot.fr)'s status on Sunday, 15-Dec-2024 06:08:01 JST Nemo_bis 🌈

   in reply to

   • Ryan Baumann

   @ryanfb This actually exists or used to exist at various torrent indexers, but it's no use when nobody seeds and the webseeds are down. There would need to be a push for everyone to use p2p protocols as primary download method, a bit like in PeerTube.

   • Embed this notice
     Ryan Baumann (ryanfb@digipres.club)'s status on Sunday, 15-Dec-2024 06:08:02 JST Ryan Baumann

     in reply to

     • raffaele

     @raffaele the CLI lets you crawl to build and update a bulk torrent/metadata collection yourself from scratch, what's needed is that entire collection continuously updated and downloadable on its own either offsite or in a decentralized way

   • Embed this notice
     raffaele (raffaele@digipres.club)'s status on Sunday, 15-Dec-2024 06:08:03 JST raffaele

     in reply to

     • Ryan Baumann

     @ryanfb the IA cli actually allows with ease to get bulk metadata and torrent files. But, the api called, is on the same domain, so down now.
     They offer torrent for single items, would be better to have it for full collections. The good old rsyncd or zsync http://zsync.moria.org.uk/ could also be enough.

     (I have just read Jason Scott on X, the problem is failed frontend server)

   • Embed this notice
     Ryan Baumann (ryanfb@digipres.club)'s status on Sunday, 15-Dec-2024 06:08:04 JST Ryan Baumann

     No easy technical solutions to social problems BUT sure seems like the Internet Archive should be learning from these outages and trying to develop community support to de-centralize some of their services and make them more resilient to centralized failure. One potential avenue would be publishing bulk metadata downloads, especially over e.g. BitTorrent, so that the hashes from that metadata can be used to share files over BitTorrent or IPFS when official servers are down