HomeSearchPrevious dayNext day

06:43:04gfhh1 quits [Read error: Connection reset by peer]
06:44:12gfhh1 joins
06:44:34gfhh1 quits [Read error: Connection reset by peer]
06:45:04gfhh1 joins
06:58:05IDK (IDK) joins
09:44:07gfhh1 quits [Read error: Connection reset by peer]
09:44:28gfhh1 joins
10:01:08gfhh2 joins
10:03:11gfhh1 quits [Remote host closed the connection]
11:07:27IDK quits [Client Quit]
12:01:17@OrIdow6 is now known as @Buizel
12:11:40imer quits [Client Quit]
12:14:39imer (imer) joins
14:46:53@Buizel quits [Ping timeout: 258 seconds]
14:47:12OrIdow6 (OrIdow6) joins
14:47:12@ChanServ sets mode: +o OrIdow6
15:58:49imer8 (imer) joins
16:02:17imer quits [Ping timeout: 258 seconds]
16:02:17imer8 is now known as imer
17:42:29<tomodachi94>@arkiver: Nobody's saying that... where did you get that from?
17:59:01<tomodachi94>IPFS provides the infrastructure for data storage and replication. Naturally, we still need nodes (people who are willing to store the data, 'peers' or 'seeders' in torrent parlance) along with a way to tell the nodes which content to pin (or 'seed', in torrent parlance).
18:02:13<tomodachi94>@datechnoman didn't we end up crawling IA for the original IABAK implementation?
18:05:51<tomodachi94>And it also does deduplication if there ends up being multiple of the same file in IA, or even files with largely the same content. IPFS does chunk-based deduplication.
18:24:17<tomodachi94>I wrote down some ideas for IA.BAK over IPFS if anyone's interested: https://gist.github.com/tomodachi94/f8fbaea1138f69631385083bccc5f3ef
19:30:58albertlarsan68 (AlbertLarsan68) joins
19:31:51<albertlarsan68>Hi tomodachi94, I would be interested to help you on the IAoIPFS project :)
19:33:55<tomodachi94>@albertlarsan68: It's not really a project right now, it's mostly just an idea haha.
19:34:56<albertlarsan68>If you need someone to write some (python/js/Rust) code, please ask!
19:35:51<albertlarsan68>I do think it is a good thing to do, if IA goes down then many archives will also go down.
19:36:38<albertlarsan68>And making it distributed could only increase the safety of the data.
19:37:58<tomodachi94>I feel like this project can only happen if all parties (IA, people who run ArchiveTeam servers, ArchiveTeam) engage in a high level of cooperation.
19:40:46DigitalDragons (DigitalDragons) joins
19:41:22<albertlarsan68>I think that if it is possible, being able to stream directly from IA's servers to the IPFS clients (when data is asked for) would enable running as a warrior project.
19:44:11<albertlarsan68>Maybe also rely on https://collab.ipfscluster.io/ ?
19:45:26vokunal|m leaves
20:24:52<tomodachi94>Ah, that's the name of it! I remember seeing something like that, but I forgot it's name.
20:27:39<@JAA>tomodachi94: As far as I know, IA.BAK didn't fail due to a lack of management infra but rather because there simply wasn't enough volunteered storage to make a dent and mirror a significant fraction of IA. (But the project was already dead by the time I joined AT, so that might not be correct.)
20:28:21<@JAA>IIRC, it was something like high two-digit terabytes mirrored? On the order of 1‰ of IA.
20:29:43<tomodachi94>Oh, gotcha.
20:29:46<tomodachi94>@JAA: This brings up the interesting question of how much deduplication would reduce the storage space by. Probably would be something like 0.05% of all data?
20:29:57<tomodachi94>*0.05% reduction in space
20:30:06<tomodachi94>Or probably less, now that I think about it more.
20:31:34<tomodachi94>There's also Filecoin, which is a crypto built on-top of IPFS that could provide incentive for crypto peeps to be seeders, but I'd be wary of that.
20:55:54<fireonlive>yeah please no crypto T_T
21:02:45<@JAA>Yeah, there's very little to dedupe.
21:09:46<fireonlive>the topic only makes me want to ask about it more :p
21:18:14<@JAA>You do not recognise the bodies in the water.
21:30:17<fireonlive>I do not recognize the bodies in the water.
21:54:38imer quits [Ping timeout: 252 seconds]
22:04:05imer (imer) joins

HomeSearchPrevious dayNext day