Comment by runeks | Hacker News Reader

Comment by runeks | Hacker News Reader

Parent Full thread

So, Bitcoin ABC (Cash) is three blocks behind the other nodes, but no chain split has been detected. What’s up with that?

Did the block three blocks ago cause a fork? Are we just waiting for someone to mine a Bitcoin Cash block on top of this block and then there’ll be a fork?

EDIT: Here's the log from a Bitcoin ABC node I have running:

    2017-08-01 13:23:36 ERROR: AcceptBlock: bad-blk-too-small, size limits failed (code 16) (block 00000000000000000019f112ec0a9982926f1258cdcc558dd7c3b7e5dc7fa148)
    2017-08-01 13:23:36 ERROR: ProcessNewBlock: AcceptBlock FAILED
    2017-08-01 13:23:48 ERROR: AcceptBlockHeader: block 00000000000000000019f112ec0a9982926f1258cdcc558dd7c3b7e5dc7fa148 is marked invalid
    2017-08-01 13:23:48 ERROR: invalid header received
    2017-08-01 13:24:06 ERROR: AcceptBlockHeader: block 00000000000000000019f112ec0a9982926f1258cdcc558dd7c3b7e5dc7fa148 is marked invalid
    2017-08-01 13:24:06 ERROR: invalid header received
    [ ... last two lines repeat ... ]

Looks like 00000000000000000019f112ec0a9982926f1258cdcc558dd7c3b7e5dc7fa148 AKA block number 478559 was rejected by Bitcoin ABC.

EDIT2: Here's the code [1]. The block was rejected by Bitcoin ABC/Cash because it was smaller than 1 MB.

[1] https://github.com/Bitcoin-ABC/bitcoin-abc/blob/174846aaaf56...

That's normal. Bitcoin ABC is rejecting the non-BCC blocks once the median time is after 12:20 UTC. So BTC blocks after 478558 are rejected.

Now the BCC miners are actually working on producing the first BCC-only block, which will be block 478559 on the BCC chain and will be different from block 478559 on the BTC chain, hence effectively splitting the chain.

Since BCC has only a tiny fraction of BTC's hash rate (around 1% by most estimates, although it's pretty hard to evaluate) it will take a while for this block to be found. Probably hours, maybe even days.

Eventually the difficulty on the BCC chain will adjust, so even if the hash rate remains low the "blockrate" should steadily increase within the next few weeks. Also since BCC blocks can be 8 times as big they can fit more transactions in a single block, somewhat mitigating the effect.

The adjustment is by block-count not time, if they only produce a block per day it'll take ~144 days to lower the difficulty.

I think people would abandon long before that.

The difficulty recalculation in Bitcoin ABC was reduced significantly. I believe it is 6 blocks but could be wrong. At most it would take several days as opposed to several weeks.

they probably should have hardcoded a much lower dificulty for the first batch of blocks.

There's a "cheat" to avoid this situation in the technical spec[1]:

In case the MTP of the tip of the chain is 12h or more after the MTP 6 block before the tip, the proof of work target is increased by a quarter, or 25%, which corresponds to a difficulty reduction of 20%.

So as long as the BCC keeps lagging behind massively (less than 6 blocks every 12h) the difficulty is reduced by 20% after every new block.

[1] https://github.com/Bitcoin-UAHF/spec/blob/master/uahf-techni...

Thanks I didn't know of that 'cheat', it sounds from the spec though that the difficulty of the block yet to be found would be reduced but it would then go back up for the block after that to the proper-set difficultly (because there would now be a block within the 12h window).

It's computed over 6 blocks: "the MTP of the tip of the chain is 12h or more after the MTP 6 block before the tip".

So it takes 6 consecutive blocks to be found "fast enough" (that is, within a 12h window) to return to the normal difficulty algorithm.

Couldn't big miners play this by mining 6 blocks of Bitcoin Cash and then letting it rot without blocks?

Only possible if the colluding miners have almost all the hashing power to themselves (to mine 6 consecutive blocks faster than everybody else). If that's the case then those miners can do anything they want anyway, including double-spending money[1].

Furthermore if they really only wanted to slow down BCC and nothing else it still wouldn't be very efficient. They'd just postpone the scaling for the next 6 blocks and then the rule would kick in again. So they'd have to keep make sure that they're always just at the limit of 6 blocks per 12 hours... but at this point you're actually mining block and contributing to the ecosystem, so there's no problem.

[1] https://bitcoin.org/en/glossary/51-percent-attack

Ahh I thought this 20% drop per 6 blocks was only after the split block and would end after the first 6 blocks mined in 12h, would have been too easy to game that one.

Looks like though they're getting blocks now, 4 blocks in less than two hours.

Reply on news.ycombinator.com