Topic: Ultimate Bitcoin Stress Test - Monday June 22nd - 13:00 GMT - page 2. (Read 21455 times)

Such transactions should not be allowed IMO. I tested this a while ago on litecoin and surprisingly it worked. Never gave it much thought. Hopefully some r values are reused in that test so we can break it.

Edit: Heres the litecoin test http://www.litechain.info/tx/fda9aa43c77a6dc570981339f1562f2d4ee824c1cafc513adbaf97a8ec11d6be

There is a new thread today about this address that is sending numerous transactions to itself.

https://blockchain.info/address/1CD523oyvmb9QVyABc1uu9Zt9ovjPzXV7h

Could it belong to the stress testers? Is it being used to trial what their servers can handle before tomorrow, or could it be a copycat attempting his own stress test?

Not normal ... and we are sunday, too.  

https://tradeblock.com/blockchain/

Time will show, however, I doubt it.

correct address for unconfirmed : https://blockchain.info/unconfirmed-transactions

1,5Mb requiered.

So we can take bets as to what collapses first: Their servers or the network 

  Ah, it's happening ... again.
http://statoshi.info/dashboard/db/transactions

So ... new test this monday on 8:00AM ? 

Where exactly have they published the results? They wrote that the test ultimately ended without reaching the goal due to server crashes. So show me the _useful_ results.

A lot of useful information has been published around this test, like the one from Peter Todd. But nothing that is absolutely new - otherwise the blocksize discussion wouldn't go on since a long time. And Peter Todd writes:


If the testers would come up with suggestions how to solve the problem, I would happily cheer them - and I still would ask for some background information.

Instead, there is someone claiming to be a millionair who's exiting the Bitcoin community who all the sudden spends money for this test and shows some interest again. Then there is a wallet "company" with a virtual office without giving any information about who's behind it. Then they claim to be partnering with another wallet provider. And so on. I've seen too many scams in the Bitcoin world in the past to not ask some questions.

So if you think they are legit: Do you entrust them your coins? Have you opened a wallet with coinwallet.eu? Would you? Just as a reminder: They want you to send your coins to their hosted wallet but don't give you any information about who they are. So who is stupid at the end of the day? The one who gives them the coins or the one who asks who they are?

But of course: You can call me stupid for asking. This reminds me of system administrators who happily patch their systems for years without asking who delivers the patches....

I guess at certain degree, the block size will stop its fast growth. That is when most of the people who are capable of use bitcoin have all involved

Given bitcoin's no central bank control thus high short term volatility, it will less likely to be used in daily commercial activities (And many countries have developed domestic third party/mobile payment system which can do real time and instant transactions with very low transaction fee, bitcoin does not have advantage in this area)

International remittance and long term anti-inflation storage is two major usage. These usage have a less customer base and less frequent transaction (You send money to another country once a month, saving once a month), so transaction capacity should be manageable for many years to come with a 32MB block size

International trading is another potential area, but due to the high cost of international transport, it might only happen at wholesale level, large transaction volume instead of frequency

There are conspiracy theories running around that this was:
(1) To cause the market to drop, and the "testers" would profit since they had millions in short positions set. (market didn't drop).
(2) That this was done by people within the pro 1MB Raising Camp of Devs/close supporters, to show stress.
(3) That this was done by people within the anti 1MB Raising Camp of Devs/close supporters pretending to be pro, but showed higher fees work.
(4) That this was done by miners who were trying to increase fees and show how 1MB Raising isn't currently needed. (they want more fees).
(5) That this was done by miners who were attempting to manipulate the fee market to rise before blocking halving (ex. new fee 0.0005 satos).
(6) That the location that this site (coinwallet.eu) is located is associated with a virtual office that is known to be used by HYIP and Scammers.
(7) That no one has heard of this wallet (coinwallet.eu) before and they used this test to build rep and get advertising.
(8.) whatever etc etc etc

So, some people are suspicious why someone would, that everyone doesn't know,
would blow through 20btc (which I'm not sure they did) to test the network. (Their servers crashed).

At the end of the day, it doesn't really matter, but of course, with this forum and other bitcoin related websites,
people are always fighting. This is just another one of those topics now... "Who really are these guys?"

Edit: added "s" to websites

I understand your point of view, but is the nature of the community to always wanted to have that extra detail of who did what and when, the point here is a lot of people saw the difference of what this test showed in the result and during testing it has value for them and when you have value you want to give it a name.

ahh really? why the fuck in this thread are some stupids asking for names, reasons, companies and similar bullshit? they tested network. for free. they also published results for free...and now is devs. turn to analyze and implement what is necessary.

is totally irrelevant who is behind and because it cost only 2BTC for 12 hours, some rich bored smart ass can start it right now and run it until end of his life..without single announcement or thread like this one. deal with it.

I personally don't need any evidence besides my personal experience: I tried to access blockchain.info website several times, it wouldn't show up. My transactions got frozen and confirmation times took ages.

Also, any eccentric millionaire could target an attack on BTC for the sake of it with a worse result than what we say today. Any eccentric billionaire could keep the network bloated 24/7 for as long as we wants to waste money at.. that was only something controlled. When the day of a real attack comes, there will be no one to complain.
Time to do something about this.

Depends on your client.  I read somewhere (maybe upthread here, I can't recall, there's been a bunch of similar discussions on this) that the latest Multibit allows the user to set the transaction fee to anywhere between 1kSat and 10kSat per KB.

@DarkHyudrA: I think that it would be very good to see clients looking at the size of mempool when suggesting a fee and allowing a user to change that based on the importance of a quick confirmation for the transaction at hand.  If I'm eating at a restaurant and paying to a coinbase address, I don't really care if it confirms in 10 minutes or tomorrow, the restaurant is going to get paid and coinbase will eventually my money.  But if I'm buying into a poker game that starts in an hour and I need 2 confirmations to the join the game, that's a completely different situation.

Well, the actual clients don't propose fees, they only have a fixed value on them.

yes, yes of course. What I'm proposing is that in addition to the tx size, the client should also consider the size of the mempool when "proposing" a fee.

The transaction fee is totally optional. If you don't want to get enough priority and can wait for the age to be higher so it can be accepted without fees or with a low value, it's OK.

Please read: https://en.bitcoin.it/wiki/Transaction_fees

You know, a similar idea might work.

Consider instead of auto-calculating the fee and charging it, might it instead be an idea to offer up a choice of fees?

For example say you're about to submit a transaction, and you have an "estimate delivery time" ability in your client.  It could show you the current backlog, and present you with fee options based on estimated time to pickup?  Similar to many sites today that use shipping calculators to estimate your shipping cost and based upon that how long before it'll get to you.

The problem I see is that most people either won't fully understand how to estimate fees properly, or won't care to spend the time to do so.  They just want to know the basics of "how much will it cost and how long will it take" by clicking a button and being done with it.

I think the fee should be a function of the number of unconfirmed tx at the time of sending. Would this work? Clients need to be more intelligent.