Quote from: cypherdoc on July 16, 2015, 04:31:17 PM

Quote from: d4n13 on July 16, 2015, 05:27:01 AM

Quote from: knightdk on July 16, 2015, 02:15:13 AM

So a block size limit increase would fix the problem because for this to be profitable, the attacker needs to have at least half of the block be legitimate transactions and that there still be a backlog of transactions, yes? And since there isn't that much real transaction volume, raising the limit to the suggested compromise of 8 Mb would solve this issue for now since it suddenly becomes unprofitable to continue to attack until transaction volume increases.

Good post, totally get the math... all in on the theory...

Looking at the actuals, I'm not convinced that this is actively happening, though I see how you theorize it could.

Checking block shares, and the biggest fish only holds 20%

source: https://chain.so/btc

Though I will concead that it may be possible that the smaller pools are actually the larger holders running "shadow miners" that appear to be anon, but are really wholly owned subsidiaries of the bigger players.

Another problem I see is that blocks are not full by any stretch. Miners (for reasons that are their own) are leaving many blocks 2/3 empty.

source: https://tradeblock.com/blockchain/

So any miner that had less than 38% the share could simply start filling blocks to the brim. Perhaps I'm on a tangent here, but I've never understood why miners running rooms full of ASICs don't devote a few high power CPUs (useless for mining) to doing block stuffing. It seems like getblocktemplate and hashing would be non-intersecting. So pooring $$ into CPU could be independent to pooring money into hashing.

Quote from: Quantus on July 16, 2015, 04:12:31 AM

I still have questions tho, like how high can they push transaction fees and dose this mean the end of micro transactions in the long term?

If I followed, the condition they are driving is not the fee price, but rather the spam ratio. As long as they can continue to mine some percentage X of their own spam then they make a profit even if fees go to 1 satoshi. The "canaibalism" ration X would float depending on two factors... what their current share of hash rate is, and how much normal traffic there is.

but this graph from Tradeblock shows that more recent avg block sizes are just over 700kB:

yeah, the table you put up is wrong. it shows the max block size to be 776kB which is clearly wrong. we've had a consistent stream of 950kB+ come thru over the last few weeks.