51% attack

[Peter R]

Thanks Danny.  I'd say I'm 75% convinced.  The missing 25% is due to my belief that it would be both "obvious" and "independently verifiable" that a certain miner is operating an out-of-band double spend service.  If both these conditions hold, then it seems intuitive that something could be done in a decentralized manner.  Perhaps I am wrong, however.

And to me it seems intuitive that something could be done in a decentralized manner to prevent empty blocks or excluded transactions, but I could be wrong also.

LOL, but I've already convinced myself that nothing can or should be done about empty blocks.  And I guess it is really the same problem, isn't it?  So I think I'm now 90% convinced by what Danny said.  

[jonald_fyookball]

Thanks Danny.  I'd say I'm 75% convinced.  The missing 25% is due to my belief that it would be both "obvious" and "independently verifiable" that a certain miner is operating an out-of-band double spend service.  If both these conditions hold, then it seems intuitive that something could be done in a decentralized manner.  Perhaps I am wrong, however.

And to me it seems intuitive that something could be done in a decentralized manner to prevent empty blocks or excluded transactions, but I could be wrong also.

LOL, but I've already convinced myself that nothing can or should be done about empty blocks.  And I guess it is really the same problem, isn't it?  So I think I'm now 95% convinced by what Danny said. 

Empty blocks in the context of a 51% attack.  That paralyzes bitcoin.  Much worse than a double spend.

How is it the same problem?  And can you summarize/restate the essense of what he said?

[Peter R]

How is it the same problem?  And can you summarize/restate the essense of what he said?

I just mean the same problem of "how to come to consensus in a decentralized manner." 

[justusranvier]

Empty blocks in the context of a 51% attack.  That paralyzes bitcoin.  Much worse than a double spend.

Average confirmation time for transactions goes from 10 minutes to 20 minutes and that's a paralyzing disaster worse than a double spend?

[jonald_fyookball]

Empty blocks in the context of a 51% attack.  That paralyzes bitcoin.  Much worse than a double spend.

Average confirmation time for transactions goes from 10 minutes to 20 minutes and that's a paralyzing disaster worse than a double spend?

 a 51% attack can create longest chain and then be able to mine 100% of blocks, not just 51% of blocks.

[jonald_fyookball]

How is it the same problem?  And can you summarize/restate the essense of what he said?

I just mean the same problem of "how to come to consensus in a decentralized manner." 

We are both looking to add protocol rules?

[Peter R]

How is it the same problem?  And can you summarize/restate the essense of what he said?

I just mean the same problem of "how to come to consensus in a decentralized manner." 

We are both looking to add protocol rules?

Yes, I think so.  For the Mike H idea #1 to work, the majority of the hash power must agree (otherwise you will lose money as Danny pointed out).  If the majority of the hash power agrees, then it's effectively a rule that's been added on top of the Satoshi protocol. 

I can see why extra rules are probably a bad thing, but it's not completely clear to me why they are necessarily a bad thing.  I guess perhaps since you are at risk of 51% attack no matter what the rules are, why risk a loss of confidence from added complexity?

[Brangdon]

Here are Mike's suggestions:

   1. Dishonest blocks can be identified out of band, by having honest
   miners submit double spends against themselves to the service anonymously
   using a separate tool. When their own double spend appears they know the
   block is bad.

This only works if the service accepts double-spends from everyone. I'd have thought the service was more likely to restrict double-spends to favoured individuals. For example, it could require some pattern in transaction outputs, so that only people who knew how to create that pattern could make double-spends. That knowledge could be traded on underground bulletin boards, rather like how credit card numbers are traded today.

The honest miners could attempt to figure out and replicate the patterns, and then we get an arms race with the patterns getting ever more subtle.

I'm not saying it's a bad idea, necessarily, but it's far from a panacea.

[Peter R]

Here are Mike's suggestions:

  1. Dishonest blocks can be identified out of band, by having honest
   miners submit double spends against themselves to the service anonymously
   using a separate tool. When their own double spend appears they know the
   block is bad.

This only works if the service accepts double-spends from everyone. I'd have thought the service was more likely to restrict double-spends to favoured individuals. For example, it could require some pattern in transaction outputs, so that only people who knew how to create that pattern could make double-spends. That knowledge could be traded on underground bulletin boards, rather like how credit card numbers are traded today.

The honest miners could attempt to figure out and replicate the patterns, and then we get an arms race with the patterns getting ever more subtle.

I'm not saying it's a bad idea, necessarily, but it's far from a panacea.

I agree.  I no longer think this idea is workable.  It will create more problems than it solves.

In fact, it is ironic: it would add further heterogeneity [in block acceptance criteria] as a "solution" for the problems originally caused by heterogeneity in certain miners' TX mempool acceptance criteria:  

https://bitcointalk.org/index.php?topic=585511.msg6407793#msg6407793
https://bitcointalk.org/index.php?topic=585511.msg6407938#msg6407938