bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
Yihen has joined #bitcoin-core-dev
bitdex has joined #bitcoin-core-dev
AaronvanW has joined #bitcoin-core-dev
Guest3282 has joined #bitcoin-core-dev
Guest3282 has quit [Client Quit]
cmirror has quit [Remote host closed the connection]
cmirror has joined #bitcoin-core-dev
AaronvanW has quit [Remote host closed the connection]
gnaf has joined #bitcoin-core-dev
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] hebasto opened pull request #22899: ci: Build and cache static Qt instead of downloading a pre-built one (master...210906-jom) https://github.com/bitcoin/bitcoin/pull/22899
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
Guest74 has joined #bitcoin-core-dev
Guest74 has quit [Client Quit]
baldur has quit [Ping timeout: 252 seconds]
Yihen has quit [Remote host closed the connection]
<prayank>
1. Is IPFS currently used by bitcoincore website? Or some volunteers have created mirrors?
<prayank>
2. Is it using GitHub pages right now?
<prayank>
3. What are other IPFS alternatives that were considered or you know
tralfaz has quit [Quit: Leaving]
<laanwj>
1) no, though there may be unofficial mirrors, but it was unclear how to do it without making it easy for someone to create a fake mirror with compromised binaries 2) no, it's hosted on a specialized server 3) the torrent download might be enough in practice, there isn't that much on the rest of the website
<prayank>
laanwj: Thanks
<laanwj>
and there's the tor .onion address that can be used to access it evading censorship
<laanwj>
so maybe the IPFS thing is unnecessary, i mainly created the issue at the time to create discussion because people were asking about it at the time
<prayank>
Makes sense. IPFS isn't required. Onion website and torrents are good enough for Bitcoin Core. I was researching about IPFS alternatives for one of my project. Came across this issue so wanted to confirm few things.
Talkless has joined #bitcoin-core-dev
lightlike has joined #bitcoin-core-dev
davterra has joined #bitcoin-core-dev
luke-jr has quit [Ping timeout: 252 seconds]
mxcl has joined #bitcoin-core-dev
<dhruv>
Quantum computing and it's impact on cryptography often comes up in Bitcoin conversations. Having no background, I was often intimidated/confused. I recently studied the basics and posted my notes here in case others find it useful: https://voluntarymind.com/bitcoin/quantum-computing-and-post-quantum-crypto/
<dhruv>
If you spot mistakes, please let me know and I will learn and make corrections.
<michaelfolkson>
dhruv: Cool, thanks. Thinking through how "signing a new public PQC key with the old private key to reassign ownership to the PQC public key" would work in Bitcoin script
<sipa>
michaelfolkson: literally moving your coins to a new PQC-based script
<sipa>
:)
<michaelfolkson>
New CHECKSIG opcode? e.g. CHECKQUANTUMSIG? You wouldn't be able to extend CHECKSIG...?
<sipa>
that question will be the least of our worries
<sipa>
s/will/would/
<michaelfolkson>
Haha
<dhruv>
Yeah i was just guessing there. In general my takeaway was that for now, encryption upgrades are more important than signature upgrades for most applications for quantum-resistance.
<dhruv>
More _urgent_, rather than important.
<sipa>
well, yes, privacy is forever
<sipa>
authentication only works in the past
<sipa>
but also context matters... for something like bitcoin transactions, switching to a PQC scheme comes at a tremendous cost (PQC keys/sigs are all many times larger, and less featureful)
<michaelfolkson>
I'm sure most individuals would rather lose their encryption than lose their Bitcoin but I get the 70 year thing
<sipa>
michaelfolkson: not if they've had a chance to move their coins
<michaelfolkson>
Sure, of course
<dhruv>
As we get closer to QC, off-the-shelf hardware to run nodes will also improve and perhaps there is a window of time where switching to PQC sigs is cheap enough and QC isn't widely available.
<sipa>
yes
<michaelfolkson>
In principle (in long term) there's nothing preventing QC MuSig, FROST etc right? Just haven't been discovered/designed yet?
<sipa>
right, as far as i know
<dhruv>
I don't know if any of the NIST finalist lattice techniques are additive/composable like Schnorr. Would that be the key criterion to look for?
<sipa>
they almost certainly aren't
<prayank>
dhruv: I don't understand quantum computing and related things but this post looks interesting: https://voluntarymind.com/bitcoin/2021/ (DNS seeds)
<dhruv>
prayank: That post is a few months old and things have changed. We have discovered good reasons to not authenticate DNS seeds for now and a re-work of BIP324 is WIP to allow for e2e encryption and authentication.
<dhruv>
Once we have p2p authentication (which gives us seednode authentication), we can re-think DNS seeder authentication. But for now, baking in privileged public keys into the code base seems like a bad idea given the history of the alert key.
<prayank>
Solution could have been using IP addresses if every domain used for seeds had only few IPs that never change. But looking at some of domains and A records it seems every domain has lot of IPs. Example: seed.bitcoin.sipa.be has more than 700 IP address in A records
<sipa>
prayank: that is how DNS seeds work
<sipa>
they are DNS names that resolve to the IPs of good nodes
<sipa>
the set of IPs you get back changes all the time
<prayank>
Okay. So can we use an IP address which could resolve to all these?
<sipa>
IPs don't resolve; names do
<prayank>
I was assuming this happens: domain -> one of the IPs from A records -> This node responds with thousands of IPs in a p2p message and disconnect
<sipa>
that's how -seednode works
<sipa>
which is the bootstrapping mechanism used for tor nodes
<sipa>
but if you have unproxied access to the internet, we just resolve the seed name, and enter the results in addrman
<sipa>
with -seednode, there is an additional indirection step
<prayank>
Thanks. Will read more about it.
<michaelfolkson>
Just to check I understand this additional indirection step. DNS seeds: domain maps to one IP address, this IP address provides you with multiple node IP addresses. Seednode: domain maps to multiple possible IP addresses, one of them will provide you with multiple node IP addresses?
<dhruv>
michaelfolkson: Unproxied Internet access + DNS seeds => One domain name resolves to many IPs that go into addrman.
<dhruv>
Proxied internet access + DNS seeds => One domain name resolves to one IP that acts like a -seednode.
<dhruv>
-seednode: a peer node that is only asked to respoond to GETADDR
<dhruv>
I think...
<michaelfolkson>
Ohhhh thanks, did not know this
<sipa>
dhruv: exactly
Talkless has quit [Quit: Konversation terminated!]
bomb-on has joined #bitcoin-core-dev
kexkey has quit [Ping timeout: 252 seconds]
kexkey has joined #bitcoin-core-dev
dermoth has quit [Remote host closed the connection]
jesseposner_ has joined #bitcoin-core-dev
jesseposner has quit [Ping timeout: 240 seconds]
jesseposner has joined #bitcoin-core-dev
jesseposner_ has quit [Ping timeout: 252 seconds]
Yihen has quit [Remote host closed the connection]
prayank has quit [Quit: irc thread exit]
luke-jr has joined #bitcoin-core-dev
smartin has quit [Ping timeout: 256 seconds]
keymone has quit [Quit: Client closed]
raj has joined #bitcoin-core-dev
raj has quit [Quit: Leaving]
dviola has quit [Quit: WeeChat 3.2]
jarthur has joined #bitcoin-core-dev
ArctVaulMarsHMPJ has quit [Ping timeout: 252 seconds]