Topic: A scaled up spam experiment : #SpamTheBlockchain As A Service - page 3. (Read 11044 times)

And if the infrastructure isn't ready for it and is damaged as a result, how would you fix it? I could be misunderstanding something but it seems like inoculating someone with a disease in order to study how he will react in real conditions.

That makes this idea seem like an insanely expensive experiment!  How could this ever be a feasible thing to pull off if dexX7 is right about these costs?

I mean that I can send useless txs. I am not ready to make a website

Very true, I've done back of the envelop calculation also, and it would get very expensive.
Note that we already have approx 3 tx per seconds (400KB block average), for reaching 14, it is "only" 11 tx. (or 5 tx to be just above)

Last 2 spam days experiences was done by 2 persons I thought having read somewhere [citation needed here, I may be wrong].
You are right though, estimatedfees is too high, and I will need to find a compromise about minrelaytx and estimatedfee. (Or ask the spam to customize the fees himself)

Not sure what you are talking about. Have I hurted you because I used the terms "ecosystem", "we" or "community" ? I'm not interested in the terminology, replace by what you want. I'm interested with people who wants to learn from the experience, or devs strongly opposed for specific technical reasons.

Well, tweaking the fees will be easy anyway when the website is on.
Do you want to help ? I think the most important stuff to provide is all the raw data and charts for others to analyze.
I think the spamming site (which I can take care of, even if I would appreciate help for the front end "spammer" UX) will be decoupled from charts sites, I hope to reference websites of other developers providing and presenting the data.

The value will be more in the data we can collect from this than into the spam service itself. (without great feedback, there will be no great incentives to spam)
I did not thought other devs would help, I can create a group on github if it helps where we can put all projects related to the spam service, if I'm not the only coder.

Would you like to hire me? 
Let us start 0.00005 per transaction or 0.0001 per kilobyte.

Just to put it in some perspective:

With a size of roughly 250-350 kB per transaction, that's about 4000-6000 BTC spent in total for fees, with estimatefee 3 = 0.00044247 BTC/kB, and 14 tx/s for one month.

I told in the post why I think it is necessary to do that.

What am I trying to show ? Objective data + a peak about how miners / payment providers / wallet providers / users will behave in real economic condition.
I don't want to proove anything, I hope that discussion about data will replace the "pies in the skies" discussions about what will happen.

Personally, I am curious about what will happen if UTXO set grow several order of magniture.
And also what will happen to any in memory mempool implementation will behave when the tx rate is higher than 7 per sec for long period of time.

I am asking you the data you would like to see, so anybody can learn something out of it that might change your view, and hopefully reach a common agreement.

The debate about block size is not anymore in a productive state (but a destructive one), I hope changing the situation.


I did not start coding it, but if I start I should do it in 1 or 2 weeks. I hope laurentmt, who is better than me for presenting information the right way, will help me (or provide his own separate analytics website) so we can get fixed faster.
There is also an unknown : Maybe the community will not be able to fund enough fees for spamming for a long enough time.

My goal is at least to maintain something like 14 tx/s in total tx on the network for a month. (with estimated fees for 3 block confirmation)
Maybe it won't be possible, and I will adjust fees. Or maybe I'll just ask the spammer what amount he wants to spend per spammed KB.

I'm intrigued what this might mean. I agree it's worthless debating anymore, each side is drawing on hypothetical's that are largely unknown, whereas there's only one thing that will happen for sure if we run this experiment - the blocks will be full.

Are you close to having it running?

ability?

@OP
Could you explain what you are trying to show?

What should I do to be a part of your sect bitcoin? Let me in!

TALKING ABOUT BITCOIN, EVEN IF IN A GROUP, DOES NOT MAKE YOU PART OF BITCOIN.

Some metrics I'd like to see, along with the current "spam level" and the health of major services : number of nodes active, consistency of mempool across all nodes, size of blocks mined, time between blocks mined

1) spam kbytes per day
2) miner fees of spam txs per day

As I've said above - it is easy enough to create 1x99kb spam transaction instead of 200x500 byte txs.
Anyone will be able to put his favorite photo in jpeg to blockchain for only 0.01 btc 

Frankly, I don't know. Imagine Bitpay Coinbase and a big china's miner say "FUCK IT GO XT", what would happen ? Maybe it is not possible, frankly I don't know. But being divided on an issue is hurting us, and I think better data and observation will make people's mind.
I'm in Bitcoin only since last year, it is the first big disagreement I have seen.

The goal is not to kill the network, I will shutdown the service if any big problem.

I don't know. What if it cause decentralization to drop dead because of UTXO set size ? would you continue to defend the block limit ? (again, please I don't want to debate that, just to give an example)
Maybe it will break nodes count from 6K to 1K. (In which case, I can fix the problem by closing the opened utxo, so we all know that the block limit is not the main factor of decentralization)

Maybe nothing will happen and the fees will naturally rise, killing the spam.

I don't want to speak about "this I think will happen", the spam is meant to be massive for months so we can feel the impact. (if enough funding)

I think we will learn a lot, and have time to fixup things before it becomes real.
I wonder if miners will make contracts with payment providers to process "premium" transactions. (which would be a new revenue stream for miners, mainly the big one)

IF I DO IT, WHAT TYPE OF CHART WOULD YOU LIKE ?

What kind of data would you think help to make our minds ?

I don't want to debate whether it is useful. Vote will be made with wallets. I want to know if there is any strong opposition and deep concern about that, and if it is, why.

It's good to stress the chain, but it should be ramped up gradually so as to give people time to deal with the changes for the best adaptation. You could say the sudden shock would be a better stress test, but this isn't so much a test as much as a way of incentivizing optimization. (Right?)

One other thing to keep in mind is that if this were actual actual transaction demand miners and others could figure into their future business models that they can rely on these fees and they may start adjusting their fees to try to maximize revenue (making this attack prohibitively expensive, as we want it to be). If it's just a temporary attack, even if it continues for weeks, miners aren't going to have any incentive to change things. Well, they may have a personal or altruistic incentive to raise their fees for the good of the network, but not really an economic one. And even if they raise their fees, they may not have an incentive to make it easier for users to know how much to pay.

Also, it should be noted that if this jams up the network it doesn't necessarily support either side of the debate. It's not going to force a blocksize increase if everyone is experiencing delays; hopefully it will force better resource usage through the price system, though.

The only bitcoin buisnesses are mining and exchanging.
All other buisnesses are based on fiat currency.
Hodling bitcoins - is not a buisness. Today you hodl 1 btc. Tomorrow you will hodl 1 bitcoin and 1 gavincoin. 

I think quite the opposite will happen. The majority will go with Bitcoin Core, big companies and miners will not chose to be political, they will not risk being political in order to force a change. The nuance is, if Bitcoin Core chose 20MB blocks, everyone would upgrade, but if Bitcoin Core remains with 1MB blocks and Bitcoin XT attempts to hard fork, people wont join Bitcoin-XT.

Think about it from a business owner's perspective. Would you support the de facto official version which has dozens of developers, or a new fork that has 2?

I think that is an exaggeration. 

I guess that fewer than 1% of the bitcoiners (2'000 in 200'000) are so opposed to the very idea of a hard fork -- even one with a single outgoing branch -- that they will rather use the "old bitcoin" only among themselves, with a ridiculous block rate, than user the "new bitcoin".

The other 99% of bitcoiners do not know about the issue, do not care enough to bother, are in favor of the change, or will assent to it because they do not want problems.

For the large players that matter -- large miners, exchanges, payment processors -- the 20 MB limit will be only a small annoyance. They will go with the crowd.

Thus, i expect that most players will upgrade to the large-block version, well before the fork.  They will not even notice when the fork happens, and the price will not suffer any permanent drop.

I also expect that most of those "rebels" will then give up and upgrade too, in order to recover the value of their investment.

Thus, if the "old bitcoin" does not die immediately after the fork block, it will probably die very soon after that.

one can hope that cold hard facts and statistics will help taking an educated decision