care to explain the 'stress test'? | Bitcointalksearch.org

RustyNomad

sr. member

Activity: 336

Merit: 250

Quote from: monsanto on September 10, 2015, 02:46:22 PM

Quote from: Lauda on September 10, 2015, 01:59:05 PM

Basically because of the legal pressures Coinwallet stopped their plans and did not conduct the 'stress test'. However, by making so many keys private they are essentially making users do this for them.

I'm sceptical of the legal pressures. Maybe this was the plan all along. What laws would they be breaking?

I'm thinking the same thing. Clearly some planning went into this and this whole 'stepping back from the attack' is nothing but a smokescreen.

I just hope all those guys with their scripts and bots get these transactions claimed as quickly as possible so that we can start working down that backlog.

AmDD

legendary

Activity: 1027

Merit: 1005

Quote from: Lauda on September 10, 2015, 03:05:56 PM

Quote from: AmDD on September 10, 2015, 03:01:39 PM

I understand the bigger block thing, was just curious how it would be counted. That being said, the stress comes from people "picking apart" the 53,000 transactions (via coin control) trying to piece together enough to be worthwhile as well as stay under the current block size limit. Is that correct?

I do need to read up on inputs/outputs. I never dedicated enough time to fully understanding it.

thanks for the info

That's the people that have medium knowledge (including me, while testing). Basically a transaction via Core (I'm not even going to talk about "lite" clients such as Electrum) could not automatically be created. It usually results with a error. So to avoid this one can manually use coin control, however this is very inefficient. There are people running bots and various scripts with bitcoind that should make this much faster. You are correct.

However, if you have 1000 people that make one 0.0001 BTC transaction each (total 0.1 BTC), this is going to count as 1000 transactions. The total number of inputs would remain the same (if we assume that they are all the same on those addresses).

Great, I understand. Thank you! Cool

I'll readup on in/outputs now and try to wrap my head around fully.

Lauda

legendary

Activity: 2674

Merit: 2965

Terminated.

Quote from: AmDD on September 10, 2015, 03:01:39 PM

I understand the bigger block thing, was just curious how it would be counted. That being said, the stress comes from people "picking apart" the 53,000 transactions (via coin control) trying to piece together enough to be worthwhile as well as stay under the current block size limit. Is that correct?

I do need to read up on inputs/outputs. I never dedicated enough time to fully understanding it.

thanks for the info

That's the people that have medium knowledge (including me, while testing). Basically a transaction via Core (I'm not even going to talk about "lite" clients such as Electrum) could not automatically be created. It usually results with a error. So to avoid this one can manually use coin control, however this is very inefficient. There are people running bots and various scripts with bitcoind that should make this much faster. You are correct.

However, if you have 1000 people that make one 0.0001 BTC transaction each (total 0.1 BTC), this is going to count as 1000 transactions. The total number of inputs would remain the same (if we assume that they are all the same on those addresses).

AmDD

legendary

Activity: 1027

Merit: 1005

Quote from: Lauda on September 10, 2015, 02:43:31 PM

Quote from: AmDD on September 10, 2015, 02:38:59 PM

Ok so, say it was possible to send that as a single transaction (the ~.53 BTC), would that count as 1 transaction or as 53,000 transactions?

It would be possible if we had bigger blocks, maybe 10 MB or higher. However, right now it is not possible. It would count as a single TX (i.e. transaction) with ~53,000 inputs. You need to read up on inputs and outputs of a transaction to learn more. I suggest locking this thread now to avoid unnecessary replies.

PM me if you have any more (or other) questions.

I understand the bigger block thing, was just curious how it would be counted. That being said, the stress comes from people "picking apart" the 53,000 transactions (via coin control) trying to piece together enough to be worthwhile as well as stay under the current block size limit. Is that correct?

I do need to read up on inputs/outputs. I never dedicated enough time to fully understanding it.

thanks for the info

Lauda

legendary

Activity: 2674

Merit: 2965

Terminated.

Quote from: monsanto on September 10, 2015, 02:46:22 PM

I'm sceptical of the legal pressures. Maybe this was the plan all along. What laws would they be breaking?

Don't be. Read this article.

Quote

"I’m inclined to say there is a good argument that the CoinWallet stress tests would be a breach of section 3 of the CMA 1990," Jankelewitz said.

Quote

"However," said Jankelewitz, "The courts have previously said there is a limit to the amount of ‘authorised acts’ that are acceptable before the acts become unauthorised."

He illustrated his point using Distributed Denial of Service (DDoS) attacks as an analogy, explaining that these attacks typically involve permitted acts, but the affected service provider does not authorise those acts to be carried out so many times that their service becomes overloaded.

The analogy is actually almost perfect. This is exactly what happened the last time, and what would have happened (but they backed down). This could have been their backup plan or they figured it out after these pressures started appearing. Looking at what they're doing right now, technically nobody can blame them. They aren't doing anything aside from giving away private keys that contain coins. Technically one can't sue them (I think).

monsanto

legendary

Activity: 1241

Merit: 1005

..like bright metal on a sullen ground.

Quote from: Lauda on September 10, 2015, 01:59:05 PM

Basically because of the legal pressures Coinwallet stopped their plans and did not conduct the 'stress test'. However, by making so many keys private they are essentially making users do this for them.

I'm sceptical of the legal pressures. Maybe this was the plan all along. What laws would they be breaking?

Lauda

legendary

Activity: 2674

Merit: 2965

Terminated.

Quote from: AmDD on September 10, 2015, 02:38:59 PM

Ok so, say it was possible to send that as a single transaction (the ~.53 BTC), would that count as 1 transaction or as 53,000 transactions?

It would be possible if we had bigger blocks, maybe 10 MB or higher. However, right now it is not possible. It would count as a single TX (i.e. transaction) with ~53,000 inputs. You need to read up on inputs and outputs of a transaction to learn more. I suggest locking this thread now to avoid unnecessary replies.

PM me if you have any more (or other) questions.

AmDD

legendary

Activity: 1027

Merit: 1005

Quote from: Lauda on September 10, 2015, 01:59:05 PM

Quote from: AmDD on September 10, 2015, 01:52:50 PM

The ~~public~~ private keys to said address are now being made public and lots of people are all trying to import the keys and send the coins to their own addresses. Correct? How does this spam the network? is it by the sheer number of people trying to spend the same coins or is it by the number of transactions in each private key?

Okay I have corrected the first part for you. The problem isn't really in the number of people that are trying to spend the same coins, because transactions will get the 'conflicted' status (at least they did for me). It is because there are a LOT of inputs in the transactions that they're creating. Basically there was an address that had ~0.53 BTC on it. It had around 53 000 inputs, and to send all of that a transaction of around 7-8 MB is needed (or plethora of smaller ones).
Basically because of the legal pressures Coinwallet stopped their plans and did not conduct the 'stress test'. However, by making so many keys private they are essentially making users do this for them.

Ok so, say it was possible to send that as a single transaction (the ~.53 BTC), would that count as 1 transaction or as 53,000 transactions?

Lauda

legendary

Activity: 2674

Merit: 2965

Terminated.

Quote from: AmDD on September 10, 2015, 01:52:50 PM

The ~~public~~ private keys to said address are now being made public and lots of people are all trying to import the keys and send the coins to their own addresses. Correct? How does this spam the network? is it by the sheer number of people trying to spend the same coins or is it by the number of transactions in each private key?

Okay I have corrected the first part for you. The problem isn't really in the number of people that are trying to spend the same coins, because transactions will get the 'conflicted' status (at least they did for me). It is because there are a LOT of inputs in the transactions that they're creating. Basically there was an address that had ~0.53 BTC on it. It had around 53 000 inputs, and to send all of that a transaction of around 7-8 MB is needed (or plethora of smaller ones).
Basically because of the legal pressures Coinwallet stopped their plans and did not conduct the 'stress test'. However, by making so many keys private they are essentially making users do this for them.

AmDD

legendary

Activity: 1027

Merit: 1005

Thread for reference: https://bitcointalksearch.org/topic/coinwalleteu-stress-test-cancelled-bitcoin-giveaway-1175321

Its my understanding that a bunch of addresses were created and thousands of tiny transactions were sent to these addresses. The public keys to said address are now being made public and lots of people are all trying to import the keys and send the coins to their own addresses. Correct? How does this spam the network? is it by the sheer number of people trying to spend the same coins or is it by the number of transactions in each private key?

Im thinking there are a few different pieces of this that all came together to allow this to happen and I dont know if I fully understand any of them.

Topic: care to explain the 'stress test'? (Read 671 times)