15 May 2022 |
vikulin | In reply to @5htpabuser:nekopolis.org that's not what i meant, i meant quic datagrams * I think quic datagramm can be used for a hole punching as well | 18:41:41 |
Salem Yaslem | In reply to @5htpabuser:nekopolis.org but anyway theres also uTP (μTP) that could be used & its also UDP so reserving udp:// for QUIC is just not cool use quic:// | 19:16:44 |
majestrate | if we are doing our own special snowflake stuff via quic we could just use ygg:// | 21:55:36 |
sleep pro | ygg5:// | 23:30:09 |
sleep pro | did someone propose private ygg virtual networks with different than 200::/7 prefix before | 23:38:37 |
sleep pro | as in configurable prefix for purpose of being able to run 2 ygg daemons side by side | 23:39:28 |
majestrate | it was some fd range if i recall, then i threw out the idea of using 200::/7 as it is a deprecated range for frame relay | 23:50:04 |
majestrate | i was going to use that one in lokinet | 23:50:26 |
sleep pro | lol | 23:50:35 |
majestrate | the 200::/7 one, but then i moved to a less nutty addressing mechanism | 23:50:48 |
majestrate | implementing a dht just to do that is a no go | 23:51:19 |
sleep pro | but i mean for private yggnets where u would also want to serve public yggnet | 23:51:25 |
sleep pro | also yeah DNS is mostly fine except when it isn't | 23:51:56 |
16 May 2022 |
ufm 🇺🇦 | It is not planned to make the "weight of the peer" in 0.5? Very lacking. | 01:22:20 |
neilalexander | In reply to @ufm:twinkle.lol It is not planned to make the "weight of the peer" in 0.5? Very lacking. This is actually a much harder problem to solve than it initially appears without making it potentially possible to introduce routing loops or other gamification | 08:12:23 |
neilalexander | We've talked a lot about it but not sure what the best approach is yet | 08:12:53 |
neilalexander | (although with the potential watermarking in v0.5ish, routing loops may be less of an issue somewhat) | 08:13:07 |
ufm 🇺🇦 | neilalexander: BGP solution not applicable (as-path prepending)? | 09:03:16 |
Jonathan | In reply to @5htpabuser:nekopolis.org did someone propose private ygg virtual networks with different than 200::/7 prefix before thats the 300::/8 range, right? | 09:24:41 |
sleep pro | 200::/7 includes 300::/8 but its not for private yggnets | 09:26:08 |
neilalexander | In reply to @ufm:twinkle.lol neilalexander: BGP solution not applicable (as-path prepending)? Yggdrasil doesn’t work that way | 09:28:35 |
sleep pro | was thinking more in lines of "i wanna do 8000::/7 for own private infra project communication but wanna serve public 200::/7 on some nodes too" | 09:28:43 |
| usatokenshash joined the room. | 10:18:35 |
yarnbuild | https://netfuture.ch/2022/05/web3-is-just-expensive-p2p/ | 14:56:16 |
vikulin | https://github.com/yggdrasil-network/crispa-android/releases/tag/v2.1.2 | 15:52:13 |
Tom | In reply to @yarnbuild:matrix.org https://netfuture.ch/2022/05/web3-is-just-expensive-p2p/ kinda off topic here, but to briefly put my opinion on web3 (as someone that has been professionally involved in crypto since 2015), web3 is a buzzword with no meaning. Its completely centralized and expensive. Its not even p2p. Steer clear. | 16:02:52 |
| JAE (DN0) changed their profile picture. | 16:13:26 |
vikulin | Is there any abstract type which can be assigned as result of net.ResolveUDPAddr nad net.ResolveTCPAddr for a single var? | 20:28:47 |
vikulin | * Is there any abstract type which can be assigned as result of net.ResolveUDPAddr and net.ResolveTCPAddr for a single var? | 20:28:55 |
vikulin | In reply to @vikulin:matrix.org Is there any abstract type which can be assigned as result of net.ResolveUDPAddr and net.ResolveTCPAddr for a single var? found https://pkg.go.dev/net#ResolveIPAddr | 20:47:26 |