Topic: Luke Jr's HARDFORK proposal debunked - page 2. (Read 2536 times)

the only reason luke JR wants a difficulty decrease is for his own selfishness and greed to let his own pool survive a couple months longer due to it being unable to compete against the big guys.
its not about the pools as a whole entire community. its not even about a fix that would last more then a month.

and everyone know that it does not take 200 minutes to make a block. (when miners are hashing blocl 425000 they are not using data(transactions) from block 424980(200 minutes ago)

im not a classic fanboy. i am not in any camp. i also have no allegiances to anyone so if anyone has spouted out wrong info i have no problem being frank about it.

but i do love how people protect friends before thinking about the code and community as a whole. and your only rebuttle is not a evidence based or a factual filled rebuttle, but an insulting ramble.

i dare you to prove that miners making a current block are hashing out data from blocks 20 blocks ago (200minutes). if you cant supply the proof dont reply. and let this topic die (easy solution to your problem)

Dude! Just fuck off for a year and learn how Bitcoin works! You're insulting a Core dev that runs his own pool, and you clearly don't know *anything* about Bitcoin.


Also, the fact that you're saying this shit about Luke with your proven level of understanding, makes it blatantly obvious that you're hanging around with the wrong crowd (let me spell that out for you: classic). Leave there, they're messing with your brain. Spend that year truly learning Bitcoin.

ok lets translate this into english

the data containing the hash of the previous block... and the ultimate solution of the new solved block that appends on after the previous hash is an average of 10 minutes. NOT 200 minutes.
EG block 425,000 contains the blockhash of 424,999 so the timer begins between when 424,999 was solved and when 425,000 is solved.
not 20 blocks ago. not 425000 blocks ago. just the time between one block and the next

now have a sit down. a cup of coffee.. maybe even have a biscuit and relax your mind. clear any thoughts about how often a miner gets paid. and instead think about the time between when a miner uses data of block 424,999 to start working on a solution for block 425,000

now with that mindset. look at my original post.

and again the average time it takes a miner to have previous block hash of 424,999 and combine it with the other data, and hash it away until they find a solution for 425,000 is an average of 8-12 minutes going by my random numbers.

now using the first table in the OP. even if A,F,I,M,N,O,R,S it would not bring the average to 16-24 minutes. because the although R no longer available to solve block 425,000.. G would have came up with the solution just 1minute 9 seconds later.

try to separate the mindset of how often a miner gets paid, and concentrate on the times of a solution that include the data of the previous hash, you know the time scale of 1 blocktime.

EG if i asked you to tell me the fastest time that block 425,011 was solved from the time it started hashing away data that included 425,010 previous block. you would tell me 10:44, made by pool P.

would you seriously reply to say that it must be 116 minutes and 30 seconds because pool P didnt solve a block from 425,000 to 425,010 so the timer must start from 425,000? i hope not

now if i told you to take away pool P and pretend it did not exist before even making the block.
also take away B,C,D,E,G,H,J,K,T (look at table 3 in the OP)
so now there is only 50% of pools left.

would you seriously tell me it would take another pool to start hashing block 425011, containing data of block425010.. atleast 21 minutes 28 seconds (double the other pool as luke JR suggests)?
would you seriously tell me it would take another pool to start hashing block 425011, containing data of block425010.. 200 minutes

by looking at the third table you would see that pool O got the solution if pool P didnt exist.
so will you tell me it took 33minutes 27 seconds because your counting the time from the last time that particular pool got paid
so will you tell me it took 200 minutes because somehow the number of pools has relevance to the randomness of nonce. within the work of just that one pool
no. because pool O got a solution in 11 minutes 4

remember the average time to make a block is not based on how often a particular miner gets paid
a block is the 10 minute average from when it starts handling the blockhash of the previous block until its got the fastest solution compared to competitors.
then the time is reset, as a clean slate.

i have no clue what black hole some of you have jumped into to think each block takes 200 minutes each.
the difficulty is based on 2016 blocks, each found at 600 seconds. (1209600 total seconds before difficulty adjustment).
the difficulty adjustment does not care who solved the blocks or how many times a particular pool solved blocks. it just cares about the 1209600 seconds to make 2016 blocks. and adjusts +/- depending if its faster or slower

because the difficulty does not care who got paid multiple times or who got paid only once a fortnight.. its not about payments. its about how long it takes for a pool to hash out a solution to a block that includes data of the last block.

the ultimate funny part is that if luke JR thinks a hardfork code can be implemented in april to be active for july THIS YEAR.. is acceptable.. then his own acceptance of a short grace period debunks the need for 16months of delay for the 2mb hard fork

Correct.  The hash is calculated using a block header as the input to the hash where the block header consists of:

• Version
• Hash of the previous block
• Merkle root
• Timestamp
• Difficulty target
• Nonce

The amount of time that a miner spends on a specific block header depends on their hash rate.

There are 4294967296 possible nonce values in the block header.  Antpool has a hash rate of 271.70 petahash/sec.  At that rate, they "use the same data" for about 0.000000015 seconds before they give up and move on to a new block.


Also, as soon as they've exhausted the nonce range, they dump that data and start afresh on a new block.  Most of the larger pools attempt to solve millions of different blocks per second.


There's also no point mining a block that you've already determined has no solution.


Nope.  They endlessly work on billions of sets of data for 200 minutes, throwing away all the unsolvable blocks until 200 minutes later they've finally found (and solved) a solvable block.

the hash is data+nonce
where the data is specific to that individual block.. miners do not use the same data for 200minutes.. as soon as they see someone else has a solution(in the 10 minute average). they dump that data and start afresh on the next block. as there is no point mining a block thats already solved by someone else.

miners do not endlessly work on one set of data for 200 minutes.

Actually, that's exactly what it means.

There is no "giving up".

Pools continuously calculate hashes until they find a valid solution, then they go back to continuously calculating hashes until they find a valid solution again.  There is never any progress towards a solution for any pool, or for the network at all.  Every hash has an equal chance of being a valid solution regardless of whether it is the first attempt or the 100 billionth attempt.

Your fundamental assumptions are wrong.

First of all there is empirical evidence which directly relates the network hashrate to the difficulty. As the hashrate has increased so has the difficulty and vice versa.

Secondly, you have the wrong assumption that blocks will be found by a single pool every ten minutes on average. That is the wrong assumption.

There are 2^256 - 1 possible solutions. A pool will only  search through 2^64 hashes for each set of transactions that it includes in a block. Note that each pool will have a different transaction set due to choosing different transaction and setting different coinbase script text.

For the sake of simplicity let's assume that each pool has the same hashrate and that there are 10 pools (pools can also be interchanged with individual miners). In, say, 1 hashes, 10*2^64 hashes will be searched, way short of the 2^256 - 1 possible hashes. After 10 minutes, 100*2*64 hashes will be searched. With whatever difficulty is set, within that set of hashes, there will be at least 1 block solution (actually there will probably 1 solution).

Now if there are suddenly 5 pools but the difficulty stays the same, 100*2^64 hashes will need to be searched but only 50*2^64 hashes will be searched. That means there are 50*2^64 hashes that haven't be searched yet. In order to have a high probability of having found the block, each pool will have to search another (1/5)*2^64 hashes which will take an extra (1/5)*(10-5)=10 minutes to search the remaining space to find the block.

tl;dr you are assuming that pools are searching from the same place, but they don't and all start and search different areas of the hash space.

it does not take a pool 200 minutes to solve a block.. wake up will you.
if each pool took 200 minutes then each block would be 200 minutes!

lets say pools make a block on average between 8-40 minutes. but out of 20 pools one of those is going to be lucky to get a solution faster then the rest.. which is the point of the table (average 10 minutes fastest solution)
say we have 3 pools
R    8:41
G    9:50
M  10:45

when R finds a block. usually G and M stop working dispose of their attempt and reset the counter to zero..
and starts on a new block. the counter WONT continue on until they get a solution before anyone else, because that block data is solved and dealt with.
its lost time.
from the network and code point of view a block is solved in an average of 10 minutes.. the time between when a pool gets paid is irrelevant.
again a block wont take 200 minutes to solve. a block on average takes 10 minutes,

i really wonder, where out of the darkest recesses of a black hole did you ever assume that it takes 200 minutes to make a block??

seriously..
all pools take 8-40 on average. but if 1 pool is lucky to solve first. the other pools give up
that means the timer is reset to zero when they start a new block.
the timer does not continue until 20 blocks are found.

if you think that the 20th pool has been working on only 1 block the whole time the other pools were working on the other 19 blocks.. then you have missed the fundamentals of mining

just because they 20th pool only got paid once in 20 blocks does not mean it has only been working on one block for the whole time

That's all irrelevant. You are misleading yourself in your 2nd sentence:


That's true. Now answer these two:

1) So with 20 (equal size) pools, how long does each pool (on average) take to mine one block? Answer: 10 * 20 = one block every 200 minutes.

2) Then 10 pools shut down, same question: How long does each pool (on average) take to mine one block?

Remember that you just said that the pool doesn't care about how much work the rest of the network does.


That's what number formats are for.

=RANDBETWEEN(minutes)&":"&RANDBETWEEN(seconds)
does not limit the results to be forced between 8-12 or sways them into that direction either..  they are 2 sets of randomness one for minutes one for seconds.. the former is not limiting or manipulating the latter.

my formulae was formatted that way for visual purposes of doing a screenshot where the random minutes and seconds are still random. but presentable in the MM:SS format

but fill free to use
=RANDBETWEEN(480,2400)
480=8 minutes
2400=40 minutes

and then notice there still a pattern of between 480 and 720 average (8-12minutes)

TL;DR:


[imagine the biggest possible facepalm GIF here]

firstly lets get the fundementals out of the way.

luke Jr wants to do a hard fork to drop the difficulty because he WRONGLY thinks blocks will take longer to mine due to events of the reward halving.

now to understand this you need to know some basics that luke jr has not factored in.

when a pool solves their block. the solution is based on the work done by just that pool. not the entire network.

its the data processed within that pool that keeps attempting different combinations until it finds a solution.
any 'work' done by pool B is not shared with pool A.. they are separate pools and their attempts are not related in any way.
the work is independant to the individual pool. and not shared. only the end solution is shared.. not the work

so say pool A solves a block in 10 minutes.
if pool B decided to give up and claim bankruptcy. it would not affect pool A's work at all, pool A would still have solved that block in 10 minutes.

also
block difficulty is not based on network hashrate. but based on the times of each block solved over  2016 blocks.. if its less than 2 weeks the difficulty changes.

so say pool A solved 2016 blocks in a row. it does not matter what the hash power of pool B, C, D, E is. because they have not solved any blocks.
also its not a guarantee that the pool with the top hashpower is going to solve every block. so again hashrate is not important.

now lets get into understanding the scenario
when 1 pools solves a block the other pools dump the attempts and move onto the next one(usually).

now in a nice world of statistics we would love it if when a block is solved the other pools continue on until they too have a solution. and then publicise the average time it takes for each pool to make a block.. rather then just knowing fastest first.(basically having 20 time statistics for each of the 20 main pools, instead of 1)

so my scenario is based on a random number of between 8 minutes to 40 minutes to represent the average times a miner would solve a block.
below is a table that has marked in green the 'fastest first' winner. but also has the average(random) times which the other pools would get a solution if they didnt ditch their attempts.

the reason for including the times is so that if we start to remove certain pools from the table we can see who would be next in line to win in a scenario where some of the pools didnt mine.. all times in the cells are random numbers between 8minutes to 40 minutes.. you too can easily try this yourself
i originally thought about doing 5-20 minutes randomness for the top 6 pools and 8-40 minutes for the remaining 14, to closer resemble pool hashpower percentages
 but instead thought 8-40 for all 20 pools would be fairer and less chance of knitpickers moaning.. because the numbers are larger and more fair to the naysayers



so above shows a healthy amount of pools making blocks 8-12 minutes.

now lets take away the pools that solve blocks often. lets pretend that these are the massive mining farms that luke jr presumes will go bankrupt due to the reward halving.

infact there were 20 pools listed. so lets take away the top 10 pools.. because luke Jr's presumption is that pools would take atleast 20 minutes to solve..if 50% was lost.... so lets lose the top 50% as the extreme scenario.. and see how it plays out



now.. this if luke Jr was right would show no miner making a block in less time then 16 minutes.. yet the table still shows 8-12 minutes.. so luke is wrong

now lets instead of thinking that its the top 10 pools that will go bankrupt.. but instead its the smaller pools that dont have a chance of solving as often even before the halving.



so in both scenarios blocks are still being made in 8-12 minutes.

now i want you to try it yourself.. and no before you even think about it, dont put in fake numbers to try to make a fake rebuttle example, be honest to yourself. so that you can see for yourself..
the formula i used is
=RANDBETWEEN(8,40)&":"&RANDBETWEEN(10,59)

8,40 is the number of minutes 8 to 40
10,59 is the number of seconds 10 to 59.  i done 10 instead of 1 purely for visual aesthetics because single digits look uglier. but you are free to do 1-59 as it makes no difference.

have fun making your own scenarios. and if you are going to be a twog and manually type in data just to fake a rebuttle. your only ultimately lying to yourself. so use honest random formula not manual numbers.

and you will hopefully come to the same conclusion i have.. that if pools drop out due to less rewards, it wont cause longer block solvings. because as i said if pool B drops out it wont affect pool A's work in any way