What has been seen can't be unseen!

[solex]

BIP100

MINERS
​

OK. So, this is the deal. The cat is out of the bag, the miners have seen BIP100 and they are not forgetting it anytime soon. Unlike BIP101 the BIP100 gives miners extra power, extra control. They like it. Really like it.

This implies that XT will not get its 75% threshold for activation, unless real market volumes ramp up very fast and the 1MB is maxed out for a period of time, causing a mass adoption to avoid a PR disaster and market crash and permanent reputational damage . Before this happens the likelihood is that either Jeff or someone else will write a BIP100 patch and either Core or XT or both will adopt it.

Thoughts...
Better that Jeff writes this patch than Core Dev who will throw the kitchen sink at it to keep the block limit constrained for as long as possible.

 

[cypherdoc]

That's really a sad scenario.

What was Jeff thinking giving miners such control. Ugh.

 

[solex]

I don't like it either.
Jeff came up with BIP100 first, as he was trying to find a compromise with Core Dev, who you will recall, had their heels dug in and were not coming up with any proposal despite Gavin's long series of articles. So it is a technical solution compromised by politics.
BIP101 is a pure technical solution, and I think this is what is best.
The problem now is that since XT was launched the miners are not going for it, they are looking at a big fat juicy carrot called BIP100 and want that instead. Because XT relies on miner voting, this is an impasse.

 

[cypherdoc]

Otoh, I think the number of XT nodes is still important, gaming aside. Miners can't really function without a wide network of supporting nodes. They could try but it would be unsafe. Imo, XT nodes represent the user voice of things and we should still work on that.

 

[awemany]

I think the whole situation is far from decided yet. I personally favor BIP101 and think it will eventually succeed. But I am patient with that eventually. I expect lots of movements by the various bigger parties in this system still. Hopefully, it will also show that the devs are just a part (though an important one) of this.

 

[Zangelbert Bingledack]

That's why I think the endgame is fork arbitrage on the exchanges, not miner voting. Investors are the ones ultimately in control, so why beat around the bush?

If miners choose wisely, investors will hold back and let the illusion that miners have the control stand. But if miners choose poorly (e.g., BIP100, especially if followed by miners' abuse of that power), investors will demand to be given the ability to sell that fork in favor of XT or other options. Exchanges will arrange this because it's a great profit opportunity and will drive new signups.

I'm a little disturbed that Gavin and Mike don't seem to get this paradigm shift yet. If they don't include a way to distinguish transactions among forks, the process won't work. They need to understand the endgame to enjoy maximum speed of adoption.

 

[cypherdoc]

That's why I think the endgame is fork arbitrage on the exchanges, not miner voting. Investors are the ones ultimately in control, so why beat around the bush?

this is not emphasized enough and certainly will happen

If miners choose wisely, investors will hold back and let the illusion that miners have the control stand. But if miners choose poorly (e.g., BIP100, especially if followed by miners' abuse of that power), investors will demand to be given the ability to sell that fork in favor of XT or other options. Exchanges will arrange this because it's a great profit opportunity and will drive new signups.

it's a huge opportunity for them as i've said before. it will be cheap to implement as it's all digital trading.

I'm a little disturbed that Gavin and Mike don't seem to get this paradigm shift yet. If they don't include a way to distinguish transactions among forks, the process won't work. They need to understand the endgame to enjoy maximum speed of adoption.

how would they do this?

 

[solex]

That's why I think the endgame is fork arbitrage on the exchanges, not miner voting. Investors are the ones ultimately in control, so why beat around the bush?

You are right that this is the ultimate form of arbitration, but I really hope it will not come to this. Here's why. The wider public will not understand what is happening, and will be bemused, confused and dismayed that a rock-solid digital alternative to gold can be split into two flavours so easily. If there can be two types then why not 67 types of Bitcoin?
This is a dangerous meme. All because a simple constant couldn't be quietly dealt with by the devs before it became a community issue, before it became a PR disaster.

If 2 types of BTC are listed on the exchanges I expect the combined price of both of them will be significantly lower than the unforked original for a long time to come.

 

[Zangelbert Bingledack]

It plays out in a matter of seconds, or maybe minutes. Imagine it: the very first bid/ask level might have the worse fork at a mere 1/3 the price of the better fork, then it just spirals from there. If you're fast enough you can make a quick 30% profit or so, so everyone is sitting there poised with itchy trigger fingers. In seconds I imagine the price of the disfavored fork is below 1% of the other, where it may float for a while like any altcoin, so the coins in the best fork will not even show a price fluctuation at all. To the average price watcher, they see a momentary split for a few seconds, then back to normal.

Plus then, with such immediate finality, the awesomeness of the Bitcoin market decision process will be shown to the world in its full glory, and I think investors will pile in.

Rather than being a cause of concern, I think this is the very best trick Bitcoin has up its sleeve. It's basically a prediction market for decision-making, the gold standard for settling scientific debates, built in automatically. Sure some may be confused at how it happens, but the speed at which it happens won't allow for much public consternation. I think the people who get it will really get it, and invest by the truckload. And once it happens once, it will just be accepted as part of how Bitcoin reaches consensus. Then any change that needs to happen can happen very quickly, providing a lot of extra confidence in Bitcoin's antifragility and decentralization.

 

[cypherdoc]

well, the hope is that it will be dealt with quickly and brutally. i think that will be the case.

 

[Zangelbert Bingledack]

how would they do this?

Have you read through this? Didn't see you in the comments section.

https://www.reddit.com/r/Bitcoin/comments/3i9eiv/how_i_learned_to_stop_worrying_and_love_the_fork/

 

[cypherdoc]

so i looked thru that thread but still not sure what you mean by having Gavin/Mike "distinguish" btwn tx's on the 2 chains. do you mean when someone spends coins on the XT chain the same tx on Core will not take place?

 

[solex]

It plays out in a matter of seconds, or maybe minutes. Imagine it: the very first bid/ask level might have the worse fork at a mere 1/3 the price of the better fork, then it just spirals from there. If you're fast enough you can make a quick 30% profit or so, so everyone is sitting there poised with itchy trigger fingers. In seconds I imagine the price of the disfavored fork is below 1% of the other, where it may float for a while like any altcoin, so the coins in the best fork will not even show a price fluctuation at all. To the average price watcher, they see a momentary split for a few seconds, then back to normal.

Plus then, with such immediate finality, the awesomeness of the Bitcoin market decision process will be shown to the world in its full glory, and I think investors will pile in.

Rather than being a cause of concern, I think this is the very best trick Bitcoin has up its sleeve. It's basically a prediction market for decision-making, the gold standard for settling scientific debates, built in automatically. Sure some may be confused at how it happens, but the speed at which it happens won't allow for much public consternation. I think the people who get it will really get it, and invest by the truckload. And once it happens once, it will just be accepted as part of how Bitcoin reaches consensus. Then any change that needs to happen can happen very quickly, providing a lot of extra confidence in Bitcoin's antifragility and decentralization.

Again you are right, but it seems like this is an ideal theoretical scenario. The last thing the exchanges want to do is list a 2nd version of Bitcoin that is tradable for only a short period of time. It would cause all sorts of confusion just listing a 2nd in the first place. What to call it? I consider the XT bitcoins to be the real BTC or XBT, let the Core ones be called CBT or some such new code that everyone can forget within minutes, however, I am sure that Peter Todd would have the opposite view. Whichever coin is called BTC by the exchanges benefits from historical inertia of the network effect. What if the exchanges can't even co-ordinate on the two BTC codes?

I have just read Meni's article and it leaves me cold.

A prudent user will consider the currencies as separate, starting from the block where the chain splits. Each will have its own balances, its own exchange rate against other currencies, its own list of places that accept it, etc. But he will need to be careful not to send currency A, and have his transaction accidentally picked up by network B, which will result in him losing his coins B without compensation from the receiver, who never requested coins B.

To do this he will need to decouple his coins into their components. The way to do this is to get some coins generated as coinbase in a post-split block (or coins that can have some trace of those) – those will function as a collapse agent, since they are valid in only one of the networks, say A. Then he will send all his coins – both pre-split UTXOs with his savings, and the post-split collapse agent – to himself. This transaction will only be vaild in network A, so in the receiving address he will have a coin (UTXO) with the total of his balance, that is strictly of currency A. Once this transaction is confirmed, his original UTXO is collapsed into currency B, since in network A they were already spent. Then he can manage his currency with two separate wallets, knowing for certain which currency he sends at any given time.

This is a 1MBer who would rather see a permanent split than main-chain scaling. What he writes here is something that only cryptogeeks would love, it would scare off 99% of normal users. He actually expects Joe and Jane Bitcoiner to think about UTXOs, manage 2 wallets, BTC-A & BTC-B and keep track of which part of the ecosystem recognises the coins from each wallet? FFS he has no idea how the real world works.

In the real world a lot of Bitcoin commerce would slow down until everyone knew which fork was prevailing. Merchants would be keen to be on the "right" side, and wary of coins that might be valueless long-term, unless they are relying on BitPay and others to insulate them from problems.

Back from madness to the point of this thread, which is that I can't see XT gaining much ground with the miners while some version of BIP100 is on the horizon. Unfortunately :-(

 

[Zangelbert Bingledack]

so i looked thru that thread but still not sure what you mean by having Gavin/Mike "distinguish" btwn tx's on the 2 chains. do you mean when someone spends coins on the XT chain the same tx on Core will not take place?

Yeah, just that.

...

I don't think Meni grasps the speed aspect. Obviously multiple interfaces would be a disastrous user experience,* but if you agree the "forkbitrage" process would be over in minutes it shouldn't be an issue.

*except of course in the highly unlikely event that the market decides we need two versions of Bitcoin (say one for settlements and one for smaller payments, or one for anonymity and one for efficient payments). In that case (which may well never come about), where the persistent split would be inevitable and a good thing, the user experience would be whatever it has to be. On blocksize I don't think there is any real possibility of this kind of split, so it should be over in minutes and, under that assumption, user experience doesn't really enter into the equation.

EDIT: As for the other issues with exchange coordination, I see those as only details to work out. This is the endgame, so I believe it will come about no matter what, whenever it becomes so painfully needed as to overcome inertia and name recognition. I think in hindsight it will look natural. It completes the market circle, with every process in Bitcoin being decided by the market, not just the price and the mining incentives.

TL;DR: Honey badger wants to be a creature of the market.

 

[cypherdoc]

@Zangelbert Bingledack

i personally want my privkeys to be duplicated on the XT fork so as to ensure i have an equal stake on whichever chain wins. "distinguishing" sounds like it puts that at risk.

 

[theZerg]

Yes priv keys would be duplicated ofc. But transactions should not be. That is we need to make changes so you can't take a txn from one fork and use it to spend coins on the other

 

[Zangelbert Bingledack]

@cypherdoc

Yeah that is an essential of course. The ledger must be duplicated in snapshot at that point in time, so that the market decision becomes purely one of which ledger-updating protocol to choose, not one of which ledger to choose.

Sound money means a sound ledger, i.e. one that is maintained no matter what changes in protocol might be adopted. No investor, having once put their capital onto the Great Ledger of Civilization, is ever left hanging. The fact that this status as WWL is self-reinforcing leads to unstoppable monetary network effects, regardless of any protocol changes or even any temporary missteps.

 

[solex]

@Peter R
Thinking aloud about BIP100.
You might get some useful background info by talking to Jeff in person.
Also, it would be interesting to know whether he would consider the following:

• BIP100 activation at 75% like BIP101.
• Using the vote median or mean (after bringing the out-of-range into range).

 

[cypherdoc]

Also removing the 32mb cap.

 

[Mengerian]

Interesting to see so any blocks indicating support for BIP 100 given that it is not actually implemented in code yet, and the details still seem to be being worked out.

What happens if BIP 100 reaches its activation threshold on the network, but a widely deployed software implementation is not yet available?