Nostr Web Client

I think it’s impractical and unaffordable because A: anyone who wants to (and has enough money) can write to your “secured” database (lol) and B: even authorized users cannot reuse the PoW.

Initially I thought his idea would make sense with reusable PoW:

Status quo: prove ownership of secret key to change controlled value

Lowery: provide prove ownership of secret to move bitcoin to change controlled value

But then I realized that’s the same thing with more steps.

So if there are no control values that make sense to put behind PoW, then the “pissing contest” argument doesn’t make sense to me either, because having nukes is very much a real threat. Hashing 95% of the network hashes and censoring might be a *slight* inconvenience to the opponent (expected block inclusion in 200 minutes). The dynamics of the game aren’t attractive at all when compared to the incentive to develop a nuclear capability.

DataNostrum 2y ago

> "I think it’s impractical and unaffordable because A: anyone who wants to (and has enough money) can write to your “secured” database (lol) and B: even authorized users cannot reuse the PoW."

I was referring to PoW in general, not Bitcoin's specifically. Of course PoW would not be the only security mechanism in place, just one of the pieces. Consider the Hashcash idea of using PoW to fight spam; an email with dedicated PoW attached has a lower likelihood of being spam because sending out large volumes of such messages would be very costly. Perhaps similar ideas can be applied to prevent some types of DDOS attacks.

> "So if there are no control values that make sense to put behind PoW, then the “pissing contest” argument doesn’t make sense to me either, because having nukes is very much a real threat."

Of course, no argument there, having nukes is a very real deterrent. Having thousands of nukes so powerful that you could never possibly make use of them without destroying yourself, is a pissing contest.

Reply to this note

Please Login to reply.

Discussion

mark tyler 2y ago

> PoW is a more general idea than just bitcoin

Bitcoin would have allowed for reusable PoW, which is better than non-reusable PoW for all cases I have imagined so far, and reusable PoW is still just storing a password safely.

I actually think he’s wrong, PoW doesn’t help against DDOS attacks at all, because the first thing the API request does is authenticate itself. In a DDOS scenario where the destination machine is checking for PoW, there is nothing stopping the attacking machines from submitting nearly infinite requests with invalid hashes. The destination machine is forced to check each one - which causes it work, which means the DDOS is succeeding.

I could be wrong here, I’ve never actually had to build protection against DDOS, this is just my lay understanding.

> the US and USSR fielded more nukes than they could have possibly used

Are you accounting for their calculations on how many they might lose before mounting a response? I’d be very surprised if the budget for literally useless nukes was approved.