search

bitcoinknots / bitcoin

Bitcoin Knots enhanced Bitcoin node/wallet software
MIT License
200 stars 60 forks source link

Increase `MAX_OP_RETURN_RELAY` default to 83 bytes to align with Bitcoin Core #53

Closed broadmode closed 5 months ago

broadmode commented 5 months ago

Revert MAX_OP_RETURN_RELAY to realign with #Bitcoin Core.

Looks like you inadvertently reduced #Bitcoin Core's data carrier size which had that side effect of censoring #PayNyms and #CoinJoin:

https://sourcegraph.com/github.com/bitcoinknots/bitcoin/-/commit/b7eb294a25378a3cb1c3db193f7b58109a321906?visible=10

Retropex commented 5 months ago

Instead of wanting to modify Bitcoin Knots why don't you use Bitcoin Core?

broadmode commented 5 months ago

Flavian: "The maximum size for OP_RETURN outputs used to be 80 bytes" Flavian: "miners who want to stay on 40 bytes (or any other value) can easily do so." Flavian: "I am proposing to increase it back to 80 bytes."

Luke: acknowledges, approves change

https://github.com/bitcoin/bitcoin/pull/5286

broadmode commented 5 months ago

Instead of wanting to modify Bitcoin Knots why don't you use Bitcoin Core?

Because Luke is not being forthright in his departure from #Bitcoin Core; saying that it's always been 40. A falsehood.

broadmode commented 5 months ago
  1. Luke says that Core devs are Knots developers; they are not.
  2. Luke says that OP_RETURN was 40; it was 80.

I'm merely establishing the above.

1ma commented 5 months ago

Instead of wanting to modify Bitcoin Knots why don't you use Bitcoin Core?

Because Luke is not being forthright in his departure from #Bitcoin Core; saying that it's always been 40. A falsehood.

It's not a falsehood.

Bitcoin Core 0.9.0 had the limit at 40 bytes. Proof: https://github.com/bitcoin/bitcoin/blob/v0.9.0/src/script.h#L26

Commentary regarding the ACK you referenced: https://twitter.com/LukeDashjr/status/1732844493198795184

broadmode commented 5 months ago

Instead of wanting to modify Bitcoin Knots why don't you use Bitcoin Core?

Because Luke is not being forthright in his departure from #Bitcoin Core; saying that it's always been 40. A falsehood.

It's not a falsehood.

Bitcoin Core 0.9.0 had the limit at 40 bytes. Proof: https://github.com/bitcoin/bitcoin/blob/v0.9.0/src/script.h#L26

Commentary regarding the ACK you referenced: https://twitter.com/LukeDashjr/status/1732844493198795184

"reduce OP_RETURN standard relay bytes to 40" in https://github.com/bitcoin/bitcoin/commit/8175c790eb12f0b0ca3197895a6d1d479b340b67

Retropex commented 5 months ago

It's part of Core not Knots.

1ma commented 5 months ago

Instead of wanting to modify Bitcoin Knots why don't you use Bitcoin Core?

Because Luke is not being forthright in his departure from #Bitcoin Core; saying that it's always been 40. A falsehood.

It's not a falsehood. Bitcoin Core 0.9.0 had the limit at 40 bytes. Proof: https://github.com/bitcoin/bitcoin/blob/v0.9.0/src/script.h#L26 Commentary regarding the ACK you referenced: https://twitter.com/LukeDashjr/status/1732844493198795184

"reduce OP_RETURN standard relay bytes to 40" in bitcoin@8175c79

The 80 limit didn't make it into the 0.9.0 release.

I'm referecing the v0.9.0 tag.

broadmode commented 5 months ago

It's part of Core not Knots.

That is the point of contention, sir.

1ma commented 5 months ago

Dude, you've been proven wrong.

Bitcoin Core had in fact the OP_RETURN limit set at 40 bytes in one of their official releases, namely v0.9.0.

Don't waste more of people's time.

broadmode commented 5 months ago

Nitter for those without shitter/x: https://nitter.net/LukeDashjr/status/1732844493198795184

broadmode commented 5 months ago

(Luke acked Florin's PR to take Core back to 80; pretty difficult to rewrite history now)

broadmode commented 5 months ago

"So after the live version with datacarrier transactions started with 40 to be "on the safe side", it was proposed to increase it to the previously intended 80."

Bitcoin Core continued with it's plan for 80 bytes; Luke "didn't care" because he controls Knots.

I wouldn't be here if Luke wasn't pretending Core was going rogue.

Retropex commented 5 months ago

Even if the code was modified it would not change the facts.

1ma commented 5 months ago

I wouldn't be here if Luke wasn't pretending Core was going rogue.

I don't know why you're still here. You came in claiming "Looks like you inadvertently reduced #Bitcoin Core's data carrier size" and you've been shown it was the other way around, it's Bitcoin Core who increased the limit to 80.

broadmode commented 5 months ago

Facts are stubborn things: https://github.com/bitcoin/bitcoin/blob/master/src/policy/policy.h#L72

Retropex commented 5 months ago

Facts are stubborn things: https://github.com/bitcoin/bitcoin/blob/master/src/policy/policy.h#L72

THIS IS THE CODE OF BITCOIN CORE NOT KNOTS.

1ma commented 5 months ago

Facts are stubborn things: https://github.com/bitcoin/bitcoin/blob/master/src/policy/policy.h#L72

This is master from Bitcoin Core, ser. You're referencing the latest version of policy.h source file, from Nov 7th 2023.

Learn to use GitHub.

broadmode commented 5 months ago

I wouldn't be here if Luke wasn't pretending Core was going rogue.

I don't know why you're still here. You came in claiming "Looks like you inadvertently reduced #Bitcoin Core's data carrier size" and you've been shown it was the other way around, it's Bitcoin Core who increased the limit to 80.

Because the consensus within the Core team was to come out with 40 and then increase to 80 if 40 didn't break anything.

With Ocean.xyz censoring #PayNyms and #Coinjoin, making it seem like it's Core who is going off the reservation is underhanded.

broadmode commented 5 months ago

Focus on shipping #stratumv2 instead of kneecapping fungibility.

image

Retropex commented 5 months ago

Because the consensus within the Core team was to come out with 40 and then increase to 80 if 40 didn't break anything.

Nodes runners have no obligation to agree with this consensus.

broadmode commented 5 months ago
  1. Ordinals are spam.
  2. Miners should be able to "kick them out".
  3. Pool operators should defer to miners using StratumV2.
  4. Ordinals will fail regardless, just like every shitcoin.
  5. Elevating Luke's Bitcoin over Core b/c some shitcoin is really bad for Bitcoin.
broadmode commented 5 months ago

Here's Luke agreeing that miners should be making this decisions directly:

...I'd prefer to have the default be randomised to incentivise miners to make the decision they're supposed to be making...

Why not randomize OP_RETURN in Knots like Luke wanted?

Retropex commented 5 months ago
  1. Ordinals are spam.

Cool we fight for the same thing.

  1. Miners should be able to "kick them out".

Cool that's what bitcoin knots does.

  1. Pool operators should defer to miners using StratumV2.

Thank you for discussing it with the people who develop StratumV2.

  1. Ordinals will fail regardless, just like every shitcoin.

Only if we act, bitcoin knots allows us to act.

  1. Elevating Luke's Bitcoin over Core b/c some shitcoin is really bad for Bitcoin.

In order not to raise it above bitcoin Core, do not use Bitcoin Knots.

broadmode commented 5 months ago

"there was a brief time where I was irritated that the size had been reduced to 40 bytes" on 11/16/2014

https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2014-November/006920.html

broadmode commented 5 months ago
  1. Ordinals will fail regardless, just like every shitcoin.

Only if we act, bitcoin knots allows us to act.

Censoring shitcoins = validating shitcoins.

They are failing the market test all by themselves.

There's no reason to usurp Core in Ocean.xyz for that.

Look at these unintended consequences with PayNyms and CoinJoin.

1ma commented 5 months ago

Broad, you're being unreasonable. This is a private fork of Bitcoin Core with custom modifications, and anyone can do that. There's no obligation to run Bitcoin Core as is, in practice it's a free market of choices.

Look at these unintended consequences with PayNyms and CoinJoin.

And you're making a fuss about absolutely nothing. Even though Knots has always had the OP_RETURN limit at 40, Ocean has signaled the intention to roll out Stratum v2, which will devolve mempool policy to their miners.

So at no point are PayNym and Whirlpool transactions at any risk of not being mined. Not by the network at large, and neither by Ocean (eventually).

broadmode commented 5 months ago

Until Ocean.xyz rolls out #StratumV2, it's just another gatekeeper.

Is it that surprising that the space is making noise after Ocean.xyz, who launched with the promise of decentralizing mining, is learned to have changed policy in #Bitcoin Core that has been standardized for nearly 10 years?

So at no point are PayNym and Whirlpool transactions at any risk of not being mined. Not by the network at large, and neither by Ocean (eventually).

I want to believe this.

kristapsk commented 5 months ago

side effect of censoring #PayNyms and #CoinJoin

It's not "censoring" coinjoins per se, Samourai is the only coinjoin implementation that uses OP_RETURN in process.

broadmode commented 5 months ago

side effect of censoring #PayNyms and #CoinJoin

It's not "censoring" coinjoins per se, Samourai is the only coinjoin implementation that uses OP_RETURN in process.

Ocean.xyz is censoring Whirlpool when those transactions can be trivially detected.

kristapsk commented 5 months ago

Ocean.xyz is censoring Whirlpool when those transactions can be trivially detected.

No, they don't, they just have limit on maximum OP_RETURN size, Whirlpool transactions are not handled in a specific way.

dzyphr commented 5 months ago

Anyone using Samurai and thinking they are private is lying to themselves and others.

broadmode commented 5 months ago

"80 bytes is standard policy on the network" -Adam Back

https://twitter.com/adam3us/status/1733409334577541162?s=20

"Samourai has no obligation of any kind to 'fix', they're using network standard..." -Adam Back

https://twitter.com/adam3us/status/1733409758508367972

"Luke often is correct, but sometimes only with a non-standard definition." -Adam Back

https://twitter.com/adam3us/status/1733098596122198306

Facts are stubborn things.

luke-jr commented 5 months ago

It was not reduced. It was always 40. Samourai chose to disregard and exceed it. If you or they want things to change, they should fix their code. There is no justification to change the default from 40.