Gold collapsing. Bitcoin UP.

[Zarathustra]

> I can't understand why anybody still thinks he has to say something meaningful at this point. He hasn't been able to express a single useful thought.

I think the network topology (mesh vs. small world near complete graph) for example wouldn't be discussed at all if CSW didn't start to talk about it again and again. @Zangelbert Bingledack repeatedly discusses the points he's bringing up. Is he someone of those who repeatedly gets bamboozled?

 

[jessquit]

As long as the "unicorn-index" does contain pumpy shitcoins, the bottom is not in.

Just had a look in the Wayback machine. Doge, Stellar, BitShares, BanxShares in top 10 throughout much of 2015, when the last bottom came in.

I think you're making a valid point but unfortunately the "rationality index" apparently still has shitcoins even at the bottom.
[doublepost=1523540724][/doublepost]I think CSW makes an ass of himself in public but his points re: small world mining are being lost amidst the hateful clamour that arises every time he speaks. I think it's true that small-world mining changes assumptions about SM and all kinds of other aspects of Bitcoin security. However people are so busy trying to put the dunce cap on Craig they can't hear this one possibly very significant point he keeps going on about, and about which I think he is making a perfectly valid point.

 

[Mengerian]

@Christoph Bergmann I think you did a good job describing your mixed feelings about the whole Craig Wright / Selfish Mining issue. I have a somewhat similar take.

I think part of the problem with the Selfish Mining debate stems from Emin Gun Sirer's initial overselling the "Bitcoin is Broken" narrative to promote his research. So people reacted emotionally to that, and are arguing the issue from an emotionally defensive position. This creates a situation where the emotions override the rational arguments. I think the actual situation is that Selfish Mining is a real thing, and should be analysed so that it can be mitigated. But is a fairly minor problem, and the mitigation measures won't be difficult (for example Bitcoin ABC devs are already working on one such measure https://reviews.bitcoinabc.org/D1196).

I'm also skeptical of the narrative being pushed that all decisions should be "scientific". This can be self-serving for academics and researchers who want to increase their own importance. This was one of the problems with Core, they created a cabal of technocrats who are supposedly such experts that everyone must defer to them.

The way it should work is that investors with "skin in the game" should ultimately be in control. If they make mistakes, or have the wrong theories, there is a mechanism to weed them out over time: they lose money. So I'm not too worried about entities like CoinGeek making irrational decisions, over time the forces of profit and loss should push them in the right direction. Ultimately, it would be nice to move towards a situation where skin-in-the-game is combined with rigorous academic-style research so that we can have the best of both worlds.

 

[Peter R]

@Mengerian

Well said!!

 

[go1111111]

I think it's true that small-world mining changes assumptions about SM and all kinds of other aspects of Bitcoin security. However people are so busy trying to put the dunce cap on Craig they can't hear this one possibly very significant point he keeps going on about, and about which I think he is making a perfectly valid point.

By "small-world mining" you just mean miners are directly connected to each other? If so, Craig could have just made the very simple point "If miners are all directly connected, it's really hard to increase gamma much above 0 via a Sybil attack." Pretty much everyone would agree with this. It's not that insightful.

This is supposed to be the big contribution that Craig made to the discussion which we should use to excuse all the false and silly things he says?

 

[jessquit]

@go1111111 I think it bears deeper consideration. Have we really considered all angles? For example, if miners are fully-connected, then they have an incentive not to hide their hashpower, but to "build reputation". This may make it harder to conceal a SM attack, and may make possible sanctions more likely.

The point is: emergent topology matters, and everything needs to be reevaluated in that light, not just SM, but everything

 

[go1111111]

@jessquit no one was really that concerned about selfish mining for a variety of reasons before CSW started attacking the paper. Most people already believed as Mengerian mentioned that Emin initially overhyped the practical effects of his paper. So CSW's claim that selfish mining isn't a huge deal in practice was also not very novel. What was novel were a bunch of false statements and bad arguments that he made to support his case.

Sure, maybe the idea that all miners are directly connected has some deep implications that we haven't considered yet, but until I see CSW or one of his technobabble-to-English translators put forth some clear explanation of what these implications are and why they're important, I don't see why listening to CSW is worthwhile.

 

[Mengerian]

Nice video about Lightning Network:

 

[cypherdoc]

Governing Dynamics gentlemen. still applies to miners when i first brought it up years ago and still applies today. none of them are going to do something in a reckless attempt to grab all the marbles/money (the blonde) out of fear of no one getting anything (getting crushed by the banks and gvts). all those nerds at the table are still at risk of the captain of the football team walking in the door and taking all the girls for himself. even Bitmain, who understands it's still just a small fish in a big pond:

 

[Peter R]

@go1111111

The mental model I've had for the network for quite some time was one where the miners were tightly connected and formed a sort-of "nucleus," around which non-mining nodes spread out in a random mesh around them. That miners had an incentive to be tightly connected had been known since we understood that orphaning due to propagation delay imposed a real cost on a miner (which goes at least as far back as this post by Gavin from early 2013).

And we've had empirical evidence that miners were much more tightly connected than non-mining nodes for a long time too. For example, here's a study from TradeBlock in 2015 that found blocks propagated to non-mining nodes at a rate of about 17 seconds per MB (before Xthin). As a comparison, James Hilliard had been (and still is) monitoring miner pool's stratum servers here, which showed that blocks (or at least block headers) spread around the world to the hash power an order of magnitude faster.

So CSW didn't really add anything new to the theory nor did he present any new empirical data to support it. One thing he did do was give it a fancy-sounding name: "near complete graph." In hindsight, I think this was effective in getting less technical people to understand what was already fairly well understood. Giving it a name gave people something to hold on to and made it more real to them. So I think it would be accurate to say that CSW popularized the concept.

All that said, and AFAIK, we still don't have good empirical data on how the hash power is actually connected or what the actual "network distance" is between mining nodes. To what extent it is really a "complete graph" is unknown to me.

 

[Richy_T]

Wasn't that what the Bitcoin Relay Network (and descendants) was/is all about anyway? It's not like this is speculative stuff.

 

[cypherdoc]

yes, we've always known it as the original "relay network" from Matt Corallo: https://bitcointalk.org/index.php?topic=766190.0, which as you can see he's upgraded to Fibre.

 

[Peter R]

@cypherdoc

Yes, indeed.

If I remember correctly, you were a bit worried about the centralizing effect of the relay network, weren't you?

 

[cypherdoc]

yeah, i probably was b/c i wasn't sure of the effects on the non mining p2p nodes which were just coming into being as pools and corporate miners were starting to grow. our thinking over the years since has since advanced in realizing that non mining nodes are not nearly as important as core would have you believe. i still think there's a place for them with merchants and exchanges though.

i remember Greg used the relay network against your first orphaning paper as well.

 

[solex]

Wasn't that what the Bitcoin Relay Network (and descendants) was/is all about anyway? It's not like this is speculative stuff.

Absolutely!

Further, there are many instances of people pointing out that the Corallo Relay Network was a centralising sub-network, and that using Xthin, and also Xpedited, was an important decentralized improvement where any miners and non-miners can configure themselves for fast block propagation:

Greg Maxwell is wrong to think it is necessary, whatever advantage a relay network can deliver, it should be made part of the default protocol, for now there is a superior solution it's called Xthin which is the ideal solution.

It was the permissioned nature of the Corallo network, with its 1MBtard gate-keeping owner, which was a major factor in XT failing to gain traction in 2015, as miners using XT needed to use slower alternate routing for block propagation.

 

[cypherdoc]

was it permissioned though? i always remember the code being open sourced and was even considering using it for my small mining operation at the time. it wasn't worth the effort though.

 

[solex]

Yes, the code was open sourced, but like Bitcoin itself, not much use to anyone else unless a critical mass of users, i.e. miners, switched over to use a different instance. With "permissioned" I mean the miners were constrained to 1MB by the relay network, independently of the bitcoin protocol.
"Limited" is a better word.

 

[Zarathustra]

@go1111111

So CSW didn't really add anything new to the theory nor did he present any new empirical data to support it. One thing he did do was give it a fancy-sounding name: "near complete graph." In hindsight, I think this was effective in getting less technical people to understand what was already fairly well understood. Giving it a name gave people something to hold on to and made it more real to them. So I think it would be accurate to say that CSW popularized the concept.

Yes, popularizing Bitcoin Cash is very important. Without charismatic individuals (Peter Rizun, Roger Ver, Craig Wright, Steve Jobs et al.) those projects would never become popular. Whether they are accused of plagiarizing or using babble speech is less important to the success of the project.

 

[Peter R]

@Zarathustra I see your point, but CSW's plagiarism is a bit more than an accusation at this point:

https://coinjournal.net/craig-wright-accused-of-plagiarism/

 

[jessquit]

@jessquitSure, maybe the idea that all miners are directly connected has some deep implications that we haven't considered yet, but until I see CSW or one of his technobabble-to-English translators put forth some clear explanation of what these implications are and why they're important, I don't see why listening to CSW is worthwhile.

TLDR "la la la la I'm not listening la la la"

Thank you for completely making my point for everyone to see.

I bring up small world mining, you bring up Craig Motherfucking Wright as some sort of hex against discussion, and tell me that you're literally not going to think about it, since he said it.

Whatever we do let's make sure to drag Craig's name back into the conversation over and over like some stinky red herring to make sure everyone keeps their fingers in their ears and doesn't discuss small world mining.

If you cannot get over your obsession with personality to focus on ideas then you add negative value to the conversation.

I don't care if Satan himself has something to add, I'm not going to stick my fingers in my ears. Since I don't evaluate ideas based on the mouth they originate from, the fact that the speaker is an asshole is irrelevant to me.

Now can we PLEASE have a conversation about the ideas without turning it into an argument about why we would all think more clearly with our fingers in our ears?

Edit: I just realized. The noted scientists among us are actually afraid to pick up the thread of small world mining, lest they get the stink of Craig on them. As soon as anyone with reputation advances any thought in this area, their acknowledgement will be used by the howling masses to paint them as "in support of Craig" and if there's anything more important than getting to the bottom of an idea, it's making sure that the popular people don't think you're friends with a loser like Craig. Distance must be maintained. As cliquish as middle school. SMH.