implications of merged mining / shared blockchain

phelix
Posts: 1634
Joined: Thu Aug 18, 2011 6:59 am

Re: implications of merged mining / shared blockchain

Post by phelix »

jtimon wrote:
phelix wrote: I have some namecoins and have registered a couple of domains and so it bothers me but bitcoinEXpress's actions are totally legit. There is a flaw and he will exploit it and even publicly announces so.
Why does he exploit the flaw when its solution is going to be launched soon?
The word "legit" doesn't sound very good in this context to me.
Mtgox and myBitcoin attacker's actions are legit too?
If his motivation is really "get data from an experimental attack" instead of stealing from others why can't he just attack a testing chain?
Can I enter in your house, burn it and say "Hey, your lock is not secure enough"?
Does the fact that I warn you first change the fact that I'm burning your house?

bitcoinEXpress: The lock of your house is not secure enough: I have machine that I think can brake it. So one of this days I'll try to enter in it and burn it all to see how much I can destroy with my machine.

phelix: Oh, thank you. That way we could learn a lot about secure locks. You even warn me, what a gentleman!
I get your point. still I consider it good he puts some pressure on the problem and he that does it transparently. He or somebody else could just have ripped off doubleC and maybe others.
nx.bit - some namecoin stats
nf.bit - shortcut to this forum

BitcoinEXpress
Posts: 15
Joined: Sat Sep 10, 2011 7:35 pm

Re: implications of merged mining / shared blockchain

Post by BitcoinEXpress »

jtimon wrote:
phelix wrote: I have some namecoins and have registered a couple of domains and so it bothers me but bitcoinEXpress's actions are totally legit. There is a flaw and he will exploit it and even publicly announces so.
Why does he exploit the flaw when its solution is going to be launched soon?
The word "legit" doesn't sound very good in this context to me.
Mtgox and myBitcoin attacker's actions are legit too?
If his motivation is really "get data from an experimental attack" instead of stealing from others why can't he just attack a testing chain?
Can I enter in your house, burn it and say "Hey, your lock is not secure enough"?
Does the fact that I warn you first change the fact that I'm burning your house?

bitcoinEXpress: The lock of your house is not secure enough: I have machine that I think can brake it. So one of this days I'll try to enter in it and burn it all to see how much I can destroy with my machine.

phelix: Oh, thank you. That way we could learn a lot about secure locks. You even warn me, what a gentleman!
Not going to steal anything, The NMC creators have about 10 days time so shame on them if I am able to do this.

jtimon
Posts: 27
Joined: Fri Jul 22, 2011 5:36 pm
os: linux

Re: implications of merged mining / shared blockchain

Post by jtimon »

BitcoinEXpress wrote: Not going to steal anything, The NMC creators have about 10 days time so shame on them if I am able to do this.
Oh, great, you're only going to destroy, not stealing anything.
What are they supposed to do? A checkpoint?
Why do you want that?

phelix
Posts: 1634
Joined: Thu Aug 18, 2011 6:59 am

Re: implications of merged mining / shared blockchain

Post by phelix »

going back 10.000 blocks is way too much to prove the point. what about going back a 100 blocks?

but a checkpoint is a good idea anyway, even if there was no attack at all.
nx.bit - some namecoin stats
nf.bit - shortcut to this forum

jtimon
Posts: 27
Joined: Fri Jul 22, 2011 5:36 pm
os: linux

Re: implications of merged mining / shared blockchain

Post by jtimon »

phelix wrote:going back 10.000 blocks is way too much to prove the point. what about going back a 100 blocks?

but a checkpoint is a good idea anyway, even if there was no attack at all.
Maybe making the change to merged mining earlier can solve the problem too?
I don't really know how these checkpoints work.

johntobey253
Posts: 17
Joined: Mon Jun 13, 2011 3:58 am
os: linux

Re: implications of merged mining / shared blockchain

Post by johntobey253 »

jtimon wrote: Maybe making the change to merged mining earlier can solve the problem too?
I don't really know how these checkpoints work.
If you run a client with a checkpoint at block 18932 (example) it will not recognise any attempt to rewrite blocks prior to 18932. This will avoid confusion, but it won't ensure the ability to function beyond block 19100.

To ensure that, we have to find out who controls more mining power: people who want Namecoin to die or people who want to merge-mine it. We must therefore court the universe of BTC miners. Many do want Namecoin to die, perhaps because they fear the effect of a superior competitor on the BTC price. I hope we can assuage such fears and avoid generating spite as befell SolidCoin.

BTC and NMC are natural allies against fiat. But like brothers, they will fight from time to time. Let's keep this wrestling match friendly and not gouge out any eyes.

jtimon
Posts: 27
Joined: Fri Jul 22, 2011 5:36 pm
os: linux

Re: implications of merged mining / shared blockchain

Post by jtimon »

I didn't thought about that. With merged mining, if most btc miners want namecoin to die, it will die.
So what they fear is namecoin being is used as a general purpose currency to compete with bitcoin and devalue it.
The point is why miners would prefer attacking another network for free instead of directly profiting from it?

Here's my proposal:
Making a checkpoint "now", another at 19099 and start merged mining at that point. The first blocks after that will not be secure until honest merged miners overcome the attackers, so people should not make transactions until people can somehow know that happened.
We would suffer a temporal paralysis, but no value will be destroyed if finally most bitcoin miners prefer to mine namecoins rather than freeze them. If that never happens, merged mining was a bad idea after all.
How Satoshi would have thought that miners would vote for prohibiting other currencies in merged mining?
The incentives are there to avoid it.
My technical question is:
Is it possible to distribute a client with a checkpoint in the future (19099)?

By the way, does anybody knows what vinced thinks about all this?

nodemaster
Posts: 172
Joined: Wed Jun 15, 2011 12:46 pm
os: linux

Re: implications of merged mining / shared blockchain

Post by nodemaster »

jtimon wrote:By the way, does anybody knows what vinced thinks about all this?
My guess: Sit back, watch this pile of headless chickens and wait until everybody concentrates. If they managed to focus try to find out the most agreed on solution and implement it. :lol:

SCNR. Don't want to offend anybody, but we should calm down a bit and focus on producing results. We have a bunch of good ideas and should agree on new binaries. I'm loosing oversight from all those potential patches and binaries. :|
Access .bit domains with Firefox in 4 easy steps: https://masterpool.eu/proxy
MasterPool Namecoin Mining Pool

johntobey253
Posts: 17
Joined: Mon Jun 13, 2011 3:58 am
os: linux

Re: implications of merged mining / shared blockchain

Post by johntobey253 »

jtimon wrote:If that never happens, merged mining was a bad idea after all.
I wouldn't go that far.
jtimon wrote:Is it possible to distribute a client with a checkpoint in the future (19099)?
Not in the way you mean.

jtimon
Posts: 27
Joined: Fri Jul 22, 2011 5:36 pm
os: linux

Re: implications of merged mining / shared blockchain

Post by jtimon »

jtimon wrote:Is it possible to distribute a client with a checkpoint in the future (19099)?
Not in the way you mean.
[/quote]
Well, then I it would be just a checkpoint "now" and merged mining before the announced attack. 19050 ?
They have claimed both, that merged mining won't happen and that they will go to the merged mining zone after rewriting the chain and be able to mine bitcoins while attacking.
What we need is to have an honest chain that has reached merged mining. So we should start merged mining earlier to reduce the effects of the attack.

Should we be talking about it in another thread?

Post Reply