The 2021 weird mining nonce: "The Race To 4B is On"

[kaggie]

Did you catch this weird mining nonce in the first half of 2021?

It was a sentence, "The race to 4b is on", although some characters were changed to 733t speak (encoded in base..64? or was it base58?).
The nonce awarded 1 bitcoin to the miner instead of the possible 6.

Mining nonces are supposed to be random, so a full sentence seems.. intentional?

It is my guess that mining rewards are not entirely random, but based on previous transactions. If this is true, this could give certain miners an advantage who saved all of their transaction details -- and determined what part of their transactions have the possibility to become parts of future mining nonces. Perhaps the sentence was intentional, somehow? Or is that implausible?

Did you see this mining nonce? I lost the full details or I'd post it, sorry.

[1715142684]

1715142684

[1715142684]

1715142684

[1715142684]

1715142684

[ranochigo]

That is untrue. You can choose to fix a specific nonce and change something else in the block header. It is just terribly inefficient in most cases. Miners are entitled to claim up to the current block rewards, which is at 6.25BTC and the transaction fees. Any miners which is building a block can choose to include all of block rewards, part of it or none of the block rewards. This has absolutely nothing to do with nonce or the choice of it.

Let me know the block height, might've missed this before.

[kaggie]

That is untrue. You can choose to fix a specified nonce and change something else in the block header. It is just terribly inefficient in most cases. Miners are entitled to claim up to the current block rewards, which is at 6.25BTC and the transaction fees. Any miners which is building a block can choose to include all of block rewards, part of it or none of the block rewards. This has absolutely nothing to do with nonce or the choice of it.

Let me know the block height, might've missed this before.

Sorry, I lost the nonce / block height. It was in the first half, probably first quarter of 2021. It was an award of 1 bitcoin, which can help separate out which of the mining rewards did this. There was absolutely this sentence within the mining nonce, whatever the cause is.

(Even if untrue, I do wonder what the effect of having previous transactions help lead to future mining nonces would be. This might be a way to balance layer 1 and layer 2 needs, as you don't want layer 1 to be completely abandoned, and there should be benefits for transacting on layer 1, while at the same time miners should benefit for keeping the systems running. Proof of transaction+work is an interesting model.)

[ranochigo]

Sorry, I lost the nonce / block height. It was in the first half, probably first quarter of 2021. It was an award of 1 bitcoin, which can help separate out which of the mining rewards did this. There was absolutely this sentence within the mining nonce, whatever the cause is.

Hmm, can't find it. The lowest I can find thus far would be 0 back in 2017. Maybe someone else can find it.

(Even if untrue, I do wonder what the effect of having previous transactions help lead to future mining nonces would be. This might be a way to balance layer 1 and layer 2 needs, as you don't want layer 1 to be completely abandoned, and there should be benefits for transacting on layer 1, while at the same time miners should benefit for keeping the systems running. Proof of transaction+work is an interesting model.)

As of now, the only reason why you can get a block is if the hash of your block header meets the target. If you want any other conditions to affect it, then there will be a hard fork and the supply would be quite difficult to balance. Layer 1 won't be abandoned, settlement is done on it.

[kaggie]

Hmm, can't find it. The lowest I can find thus far would be 0 back in 2017. Maybe someone else can find it.

It was a reward of 1.

Damn, I should have found it before bringing it up. I have the block height on a computer somewhere, but I was seeing whether anyone else noticed so I didn't have to search for it. I'll tag you if I find it. Maybe it was the hash instead of the nonce that turned into this sentence. The 733t-script sentence was in the middle of other random characters, so not exactly a easily scriptable search.

[pooya87]

Blockchair explorer can't find any coinbase transaction that had smaller than 6.25BTC output except the one with 0 from 2017:
https://blockchair.com/bitcoin/transactions?s=output_total(asc)&q=is_coinbase(true)#f=hash,block_id,output_total,time,is_coinbase

Why do you call it "nonce"? It is the coinbase script that contains the block height and any arbitrary bytes until total length is 100 bytes.
These strings don't mean much, they are as I said arbitrary. Mining pools normally put the name of their pool there.

The 733t-script sentence was in the middle of other random characters,

The starting part is the block height, when converted to a ASCII/UTF8 string it will show up as random characters. The end part is possibly the extra nonce which is again another integer which can show up as random characters.

[franky1]

nonce is only 4bytes.
the timestamp is only 4bytes whereby part of those bytes is used as an extra nonce
timestamp can only be adjusted by 70 minutes = (upto 4200 permutations) = 12bits (1.5bytes of timestamps 4bytes)

5.5bytes is not enough nonce and extra nonce to write "The Race To 4B is On"
the best you can get. assuming it uses 6bits per character (64 possible characters of lower uppercase alphabet and digits)
44bits = 7 characters = atbest 'daRace4'

let alone for ASIC miners to have been able to create a block where the nonce/extra nonce fulfills a resulting hash of acceptable difficulty AND just happens to spell out a readable sentence when said nonces are converted to human readable characters

so im calling BS.
...
however what is possible is a pool or even random user transacting can put one of their output UTXO's as not a spendable bitcoin key but instead a random phrase.

but the topic creator has not returned to show proof of this special block/transactions existence.

but definitely not possible to have the nonce of an acceptably difficult block be also a human translatable phrase of the said length the topic creator mentioned

[DannyHamilton]

however what is possible is a pool or even random user transacting can put one of their output UTXO's as not a spendable bitcoin key but instead a random phrase.

Another thing that is possible is a pool or solo miner to put a phrase in the input (extraNonce) of the generation transaction (As Satoshi did in the Genesis Block).  Perhaps this is what the OP saw? If so, it's pretty meaningless. Those messages have no effect on the operation of Bitcoin.

[franky1]

however what is possible is a pool or even random user transacting can put one of their output UTXO's as not a spendable bitcoin key but instead a random phrase.

Another thing that is possible is a pool or solo miner to put a phrase in the input (extraNonce) of the generation transaction (As Satoshi did in the Genesis Block).  Perhaps this is what the OP saw? If so, it's pretty meaningless. Those messages have no effect on the operation of Bitcoin.

there is no "extranonce" in a coinbase input.. what you mean is probably what you quoted me saying. where the output is not a valid address/signature but a space re utilised to form a message

00000050   01 01 00 00 00 01 00 00  00 00 00 00 00 00 00 00   ................
00000060   00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00   ................
00000070   00 00 00 00 00 00 FF FF  FF FF 4D 04 FF FF 00 1D   ......ÿÿÿÿM.ÿÿ..
00000080   01 04 45 54 68 65 20 54  69 6D 65 73 20 30 33 2F   ..EThe Times 03/
00000090   4A 61 6E 2F 32 30 30 39  20 43 68 61 6E 63 65 6C   Jan/2009 Chancel
000000A0   6C 6F 72 20 6F 6E 20 62  72 69 6E 6B 20 6F 66 20   lor on brink of
000000B0   73 65 63 6F 6E 64 20 62  61 69 6C 6F 75 74 20 66   second bailout f
000000C0   6F 72 20 62 61 6E 6B 73  FF FF FF FF 01 00 F2 05   or banksÿÿÿÿ..ò.

01 - number of transactions
01000000 - version
01 - input
0000000000000000000000000000000000000000000000000000000000000000FFFFFFFF - prev output
4D - script length
04FFFF001D0104455468652054696D65732030332F4A616E2F32303039204368616E63656C6C6F72206F6E206272696 E6B206F66207365636F6E64206261696C6F757420666F722062616E6B73 - scriptsig
FFFFFFFF - sequence
01 - outputs

as you can see in a coinbase transaction it needs no 'signature' because it is a reward generated coin. and so you can put anything into the signature area.. this is not a "extra nonce" this is just a script/signature area.
some people also put in extra outputs, with 0 balance given to it because they want to use the address area(public key) as a message space

"extranonce" is usually deemed the utility of changing the time stamp by upto ~70 minutes(4200 permutations(12bits of timestamps 4bytes))

those diagnosing satoshi's mining habits of 2009 to work out which blocks seemed to follow satoshis mining pattern used the time stamps 12 bits as a term for "extra nonce." and thats what it has been associated as ever since.

becasue thats what satoshi first done in 2009-2010 as his extra nonce

[DannyHamilton]

there is no "extranonce" in a coinbase input..

You know that Google is free to use, right?

https://bitcoin.stackexchange.com/questions/5048/what-is-the-extranonce

...extraNonce gets put into the input of the generation transaction...

https://en.bitcoin.it/wiki/Block_hashing_algorithm

... extraNonce field which is part of the coinbase transaction...

https://aantonop.com/glossary-archive/bitcoin-extra_nonce/

As difficulty increased, miners often cycled through all 4 billion values of the nonce without finding a block. Because the coinbase script can store between 2 and 100 bytes of data, miners started using that space as extra nonce space, allowing them to explore a much larger range of block header values to find valid blocks.

https://www.freecodecamp.org/news/how-bitcoin-mining-really-works-38563ec38c87/

...The solution to this is to add a field to the coinbase (the transaction contents of a block, stored as the merkle tree) called the extraNonce...

what you mean is probably what you quoted me saying. where the output is not a valid address/signature but a space re utilised to form a message

Nope.  That is another possibility though.  You can use something like OP_RETURN in the output script making the output unspendable, and then you can put arbitrary data into that output script. That can happen with ANY transaction.  You don't need to be a miner to do that.

as you can see in a coinbase transaction it needs no 'signature' because it is a reward generated coin. and so you can put anything into the signature area.. this is not a "extra nonce" this is just a script/signature area.

You've just finished trying to tell me (incorrectly) that the output is re-utilized to form a message, and then immediately afterwards you've proven my point by showing AND stating that the message is in the "signature area" (which is in the INPUT just like I stated). This script/signature area in the input IS used to store an extranonce when it is part of a generation transaction specifically because the generation transaction does not need a signature, so the space can contain anything (including a message). Generally when this area is used for a message the output IS a valid address, and is NOT re-utilized to form a message.  That's because this output is how the miner (or pool) gets paid!

"extranonce" is usually deemed the utility of changing the time stamp by upto ~70 minutes(4200 permutations(12bits of timestamps 4bytes))

it is not.  Changing the timestamp is changing the timestamp.  It is a value that carries meaning.  The nonce (and the extranonce) in a bitcoin block are arbitrary values that carry no meaning and exist solely for the purpose of calculating a different hash value from an otherwise identical source data.

[franky1]

the problem with using the coinbase transaction is that the POOL has to get a signal from an asic that it has tried every permutation of nonce  and so the pool then has to make a new blocktemplate with a new coinbase transaction and new hash and header to send back to the asic.

so its not the second 'variable' space on offer. its the THIRD. because an asic, without assistance from pool can use timestamp as an extranonce to add more entropy/permutations

an asic cannot change the coinbase transaction itself when its ran out of the nonce and extra nonce. so its not like a asic can just run through nonce and then continue with a different coinbase transaction without the pools involvement.

so what you find is that asics do go through all of nonce and then tweak the timestamp (without pools involvement) and then continue.. and only report back to the pool that they have ran out (once the nonce and extra nonce permutations have run out)

a coinbase output section or a signature area can be used as 3rd,4th,5th options.. but they are not option 2 which asics can handle without a pool.

when you actually look at how asics work and realise that an asic does not know the pools private keys to make new coinbase transaction. no ram or hard drive to store the blockdata to make new hash of the block to fill a block header. you soon realise mining hashes through different adjustable area's goes in 1.2.3 stage.. not 1.3 where 3 is declared the extra nonce but requires a pools involvement every 0.000X seconds per asic.

if you want to to mention all the different mundane, empty, changeable spaces that can be used to become variables. to create more permutations, fine..
but call them the option 3, 4, 4 and be sure to remember that it involves a pools action to re-supply new action 3 each time a 3rd permutation change is needed. where as an asic can do 1.2(nonce.timestamp) without a pools constant involvement at every round of changing 1.2 permutations

ill show you a good example
32bit nonce (4byte) = 4,294,967,296 hashes
                              =4294967khash
                              =4294mhash
                              =4.2ghash
by then using the 12 bits(1.5byte) of timestamp (is 4.2ghash⁴²⁰⁰)
                              =18,018ghash
                              =18thash

up until 2018 most asics did not go beyond 9-14thash..
meaning the pool requests for 'coinbase' re templating was once every couple seconds. at most.

but just imagine instead of 1-2 seconds due to 18thash of possibilities didnt exist and instead. asics requesting new coinbase tx output variables, (leading to new blockhashes leading to new templates) every 0.004thash

and where pools were dealing with hundreds of thousands of asics

would you want to run a pool that didnt bother using timestamp to give a 1-2 second breather between requests or do you want to remain adamant that timestamp is not part of any extra nonce thing and pools were rehashing blockheaders every 0.0002381 seconds for eash asic

because if you want to pretend that the coinbase transaction is part of an asics extra nonce(second option), requiring pools to play around with block header data and rehashing.. then you are trying to say that pools do 4195 hashs a second per asic
where a pool manages say 100,000asics meaning its doing 419mhash's just to manage the asics.

sorry but asics do use the timestamp.. thus bringing the bandwidth between asic and pool down and work load for the pool down to 0.5hash a seconds per asic in 2018(50khash for all 100,000 asics combined)

now which pool would you rather manage
where asics dont use timestamp as extra nonce but pool then has to do 419,000khash because uses coinbase as extra nonce
or
where asics do use timestamp as extra nonce, so pool then has to do  just 50,000khash when eventually using coinbase as extra-extra nonce
..

reality is pools dont do millions of hashes per second feeding each asics constantly. they do maybe a couple thousand hashes per asic in a ~10minute session

so would you rather feed an asic a new block header data every 1-2 seconds or every 0.0002 seconds

..
if you want to get more grammatically argumentative there are many ways to create new multiples of permutational variables in regards to making new hashes by the pool. even just changing the order of the transactions it has validated in mempool and recombine them same transactions in a different order will create a new blockheader id.
and you could if you argumentatively call all manner of options AN 'extra nonce' on top of other extra nonces.
but the initial extra nonce was to play with the timestamp as the most easiest most efficient way for asics to operate without pestering the pool every micro second

[pooya87]

~

I don't know how the pool <> miner communication is setup but what you are explaining doesn't sound realistic at all!
For example a single S7 miner can compute 4.73Th/s meaning it can go through the nonce (from 0 to max) 1100 times every second. If they increment time in header every round they end up with a timestamp that is 18 minutes in the future every second and after only 6 seconds they will end up with a block that is no longer valid.
Some people own more than one ASIC meaning with 5 or 6 of them you'll end up with an invalid block in a second.

[franky1]

~

I don't know how the pool <> miner communication is setup but what you are explaining doesn't sound realistic at all!
For example a single S7 miner can compute 4.73Th/s meaning it can go through the nonce (from 0 to max) 1100 times every second. If they increment time in header every round they end up with a timestamp that is 18 minutes in the future every second and after only 6 seconds they will end up with a block that is no longer valid.
Some people own more than one ASIC meaning with 5 or 6 of them you'll end up with an invalid block in a second.

and thats the point..
if the system was just 'nonce' + coinbase alterations.. then yes the asic will be begging the pool for new work 1100 times a second. per asic

but by using the timestamp as a extranonce an asic only has to beg just once every 5 seconds.
thats 5500x less begs for new work from the pool each time each asic..

as for the timestamp increments. they can change the time upto 70minutes of the current time(4200 seconds).
meaning 4200 more cycles of nonce. thus once per 5 seconds instead of 1100 per second

what you find is that each asic is given a different bit of work(coinbase alteration or other blockdata alteration).. they are not all going to do the same work. otherwise thats a waste of "double duty"/duplicate hashing.

so having 5 asics does not mean you pass through all the work in 1 second instead of 5. instead its 5 pieces of different work each lasting 5 seconds before each asic asks for more work from the pool

[pooya87]

they can change the time upto 70minutes of the current time(4200 seconds).

Miners are clearly not doing this otherwise we would have seen weird block timestamps all the time instead of times that are all very close to current time.