This query has been requested many instances, however not one of the solutions I’ve discovered describe a reproduceable algorithm that I can use to truly create a sound block header that hashes to an inexpensive hash worth. In actual fact, all makes an attempt produce hashes that each begin and finish with a nonzero hex digit. I’m utilizing JSON information from an S19 Bitmain Antminer linked to a mining pool. The shares usually are not solely being submitted, however accepted as proven within the JSON trade. Moreover, the net interface to the pool agrees with the outcomes. The one conclusion from that’s that these exchanges MUST lead to a sound acceptable hash which I can’t reproduce.
There are various points – the endian-ness of the info, whether or not the merkle root needs to be calculated utilizing a binary tree, or linearly. I’ve heard it’s totally different for a share than it’s for mining an precise block. Do I take advantage of the “sixth” subject, the “extranonce3” or ignore it?
I have been fortunate sufficient to discover a share that was submitted with comparatively few entries within the merkel department:
[
“070369f3b9e71bd07eba58ecee5f7769f5882a137dba7ace7aa315e546d2a355”,
“17df59c62159f5092d8eb3c67242111ce2278c60c534a100ed59dee5af251f9b”,
“e61529af81428a67d0a2a0e25d5b244ebfb67249704126526990d3284fc12a61”,
“2f783fd110ec6920f9f974a6ffe2af347906835fa590adfc9854bca1135a9f6e”,
“104233a89763f6a3d2df844d028bda3e60bfa300a0527edce1728ae208513381”,
“c4975154acf6646147041945c72792e0df34db75dfacbc3af78e3f6f903d77f1”,
“8fd8b923680fad7fd3cd87cce33261ebeeaa7ac50526e5d621cf72643cfc2c44”,
“f0ac98ac08f29c37bad7ee3506d584a39ff1856766770bfb008268600502b8d5”,
“2a74468e8ee8b02328b15097301c20f283290812007796a96ac2f30d4dc027ca”
]
now I THINK these are all at present exhibiting in big-endian, however how can we even inform?
Listed here are the opposite parameters that may be deduced from the miner/pool trade:
prevhash:”b672ef88428684ba8f9afd4f79dc17a7250204520001df9a0000000000000000″
coinbase1:”01000000010000000000000000000000000000000000000000000000000000000000000000ffffffff5a030c8a0d1d506f7765726564206279204c75786f72205465636868007105e1f5bcd6fabe6d6d8f91c003e19a377112ca3b0e19f15a5cc26ba53d696e6bd482a0693a90b79be91000000000000000000080cf”
coinbase2:”ffffffff05220200000000000017a914bf73ad4cf3a107812bad3deb310611bee49a3c7987c17cb7120000000017a914056adde53ebc396a1b3b678bb0d3a5c116ff430c870000000000000000266a24aa21a9eddddbc02a37f33c0fb04b32af121bd6892e481307b36f224f2aae103d44584fbb00000000000000002f6a2d434f524501a21cbd3caa4fe89bccd1d716c92ce4533e4d4733f459cc4ca322d298304ff163b2a360d756c5db8400000000000000002b6a2952534b424c4f434b3abfe8db1d3e3c680c9a9bf1623eb701ca3279a6a025f9d9f8e563fb12006fda0200000000″
and the remainder:
model=’20000000′;
nbits=”17028bb1″;
ntime=”67c28154″;
nonce=”16384721″;
extranonce1=’00’;
extranonce2=’ec100000000000′;
extranonce2_size=”07″;
and the mysterious sixth subject from mining.submit:
extranonce3??=’09b66000′;
I’ve efficiently been in a position to take comparable bits and items from an precise mined block and calculate the hash of the coinbase, hash out the merkel root (utilizing binary tree hash) and produce a sound blockheader. I do know I’m doing this proper, as a result of after I take the ultimate 80 byte consequence and double sha256 it, I get the very same hash that exists for the block in blockchain explorer. BUT that is with a mined block from the blockchain.
I’ve tried this with the info given above from the miner/pool trade, and I can’t provide you with a good hash. I already know that the primary merkle entry is backwards and I’ve to reverse endian it to seek out the block it corresponds to within the blockchain. The opposite entries are intermediate hashes that aren’t related to a block, so I can not confirm them, however I assume if I’ve to flip the primary, then I’ve to flip all of them. DO I embrace or exclude the sixth subject when I attempt to make the coinbase? Who is aware of?
As for the merkel tree, I’ve tried the identical binary tree hash that I take advantage of with an actual mined block, however I’ve tried the linear technique as nicely. The one a part of the 80 byte header that’s in query is the merkel root.
So, I say all that to ask, is there anybody who is aware of exactly how to do that and get a hash that’s correct? Perhaps a crytpo mining engineer that really works with these low degree particulars? I’ve spent three weeks, occurring 4, exploring the web and I can’t discover a full labored instance end-to-end exhibiting precisely how to do that.
-gt-
