1 Jun 2021 |
illeatmyhat | Maybe it assumes people want editorial control over their content, such that things they "delete" don't live a needlessly long time.
Or, because the system is eventually-consistent, that the user doesn't hold onto old stuff for too long? | 03:27:30 |
| zean gyges joined the room. | 03:27:28 |
Elpranocotro | I don't think this would change broadcast anyway, as from what I remember the DHT has a 2 hours max TTL. | 03:28:05 |
Discordian | In reply to @_discord_183309006767521792:ipfs.io long TTLs could lead to bad update propagation, If you have a TTL of 3 days, do an updates, some nodes might keep the old version for up to 3 days Yeah but if they see the update, they'll update their record accordingly, right? Do you happen to know the difference between "ttl" and "lifetime"? Been dying to know | 03:28:25 |
Elpranocotro | I think it's more the eventual-consistent side of things, you should assumes that deletion doesn't work on IPFS anyway 😄 | 03:28:48 |
Discordian | I like the idea of my record being valid for a very long time because I often think of a Mars <-> Earth scenario | 03:29:22 |
Discordian | Publishing nodes shouldn't really need to be around for a record to be valid | 03:29:43 |
Elpranocotro | mars is at most at 22 minutes radio :D, I think 3600 second is enough then | 03:29:50 |
Discordian | Wow only 22mins? For some reason I didn't grasp that | 03:30:07 |
Discordian | But still, long records, less data needs to go around, don't need it to be as reliable | 03:30:42 |
Elpranocotro | 44mins of ping is still close to what IPoAC provides | 03:30:45 |
Discordian | Nb, but has a pretty high packet loss tbh | 03:31:05 |
Elpranocotro | I think one of them has to do with cache, and the other with republication in the DHT, idk if ttl is cache or DHT idk. I would guess TTL = cache, lifetime = DHT but I'm not sure. | 03:31:18 |
Discordian | Crank both of them up and see what happens | 03:31:43 |
Discordian | I should work on ipfs-sync soon 🤔, really want a couple features. | 03:32:17 |
Discordian | Too busy though : / | 03:32:48 |
illeatmyhat | future business opportunity for a train of floating storage satellites so that nobody has to rely on UDP | 03:32:49 |
Discordian | Do it | 03:33:06 |
Elpranocotro | Outernet proposes unidirectional broadcast data, with things such as weather, linux packages and twitter 😄 | 03:33:32 |
Elpranocotro | (from space obviously) | 03:33:37 |
Elpranocotro | apparently it's now called othernet : https://othernet.is/ | 03:34:11 |
illeatmyhat | I can't find any information on how to use ipns over pubsub | 03:36:33 |
Discordian | In reply to @illeatmyhat:stirner.network I can't find any information on how to use ipns over pubsub https://github.com/ipfs/go-ipfs/blob/master/docs/experimental-features.md#ipns-pubsub | 03:37:12 |
zean gyges | so do i | 03:37:16 |
Discordian | With it enabled, you'll automatically use it when available | 03:37:30 |
Elpranocotro | https://github.com/ipfs/go-ipfs/blob/master/docs/experimental-features.md#ipns-pubsub | 03:37:31 |
zean gyges | thanx | 03:37:32 |
Elpranocotro | ah discordian was faster 😄 | 03:37:39 |
illeatmyhat | ah, but that tells me how to use it on the CLI. I'm checking out the go-ipfs CoreAPI and haven't found anything | 03:42:31 |
| expede joined the room. | 03:43:34 |