Closed: proposal to remove luke-jr from DNS seed list in bitcoin core \ stacker news ~bitdevs

pull down to refresh

Closed: proposal to remove luke-jr from DNS seed list in bitcoin core github.com/bitcoin/bitcoin/pull/29149#issuecomment-1871717493

4262 sats \ 16 comments \ @niftynei 29 Dec 2023 bitdevs

Leaving the issue at hand aside, I thought this was a good insight into how bitcoin core nodes find peers on the network (using a hardcoded list of seed nodes)

Also new to me was the list of requirements for running a dns seed node.

view all related items

4220 sats \ 3 replies \ @petertodd 29 Dec 2023

I thought this was a good insight into how bitcoin core nodes find peers on the network (using a hardcoded list of seed nodes)

Not quite correct: that's the first method they try. But if that fails, there is a backup list of IP addresses that are tried too. Also, for Tor only nodes, IIRC there's some fixed .onion addresses that are tried too. I2P probably has something similar.

If you run a node with -connect, the seed nodes and other mechanisms aren't used. Similarly, if you use -addnode, provided your node works and returns addresses, the seed nodes aren't that relevant either.

Assuming your ISP isn't themselves MITM attacking you, you only need a single "honest" peer for Bitcoin to properly connect to the P2P network. So the seed node mechanism has a lot of redundancy.

2498 sats \ 0 replies \ @petertodd 29 Dec 2023

BTW, the reason why DNS is used in the first place is because DNS is heavily cached at multiple levels. That makes it extremely difficult for the people running the DNS seeds to:

Figure out who is requesting IP addresses.
Serve different IP addresses to specific target clients.

This protects users by making it very difficult to use DNS seeds to attack people.

532 sats \ 0 replies \ @niftynei OP 29 Dec 2023

wow thanks for the great insight Peter!

232 sats \ 0 replies \ @kristapsk 29 Dec 2023 freebie

Yes, there are hardcoded list of both .onion and .i2p peers too, updated from time to time. https://github.com/bitcoin/bitcoin/tree/master/contrib/seeds

Assuming your ISP isn't themselves MITM attacking you

Running onion AND i2p would help here even if ISP is trying to MITM attack you.

3673 sats \ 3 replies \ @benthecarman 29 Dec 2023

This guy is so annoying

500 sats \ 1 reply \ @030a29f333 30 Dec 2023

They're developing a coinjoin protocol on nostr (yay?) but my God their Twitter is cancerous to look at. Always mad about something

21 sats \ 0 replies \ @benthecarman 30 Dec 2023

He's been working on it for years and last I looked it could be easily ddosed

0 sats \ 0 replies \ @02ad8239c2 29 Dec 2023

who ? the PR troll ?

682 sats \ 0 replies \ @03a5308755 29 Dec 2023

They all need to touch some grass

121 sats \ 1 reply \ @SpaceHodler 29 Dec 2023

I learned about this mechanism when I asked ChatGPT about it a few months ago.

It can be overridden manually by editing bitcoin.conf, which you have to do anyway to connect over Tor only.

0 sats \ 0 replies \ @03126e12b4 30 Dec 2023 freebie

https://jlopp.github.io/bitcoin-core-config-generator/#config=eyJfX2ludGVybmFsIjp7InBsYXRmb3JtIjoiV2luZG93cyJ9fQ==

4 sats \ 0 replies \ @0xIlmari 30 Dec 2023

Bitcoin needs more software plurality.

There should be an alternative node implementation, independent from Bitcoin Core.

There should be more decentralised solutions for things like DNS lists or node IP/onion seed lists.