[Bitcoin-ml] Transaction mining priorities.
shadders.del at gmail.com
Thu Sep 21 00:47:14 UTC 2017
On 20/09/2017 12:21 AM, Jeff Garzik via bitcoin-ml wrote:
> The miners have two notable incentives:
> 1) Maximize fee revenue
> 2) Minimize time spent creating blocks
Agree unless this is part of consensus it will only have practical
meaning until someone decides to stop being altruistic and maximise
revenue. At this point the remaining miners have a competitive
disadvantage and if margins are tight enough they may have no choice but
to adopt the same behavior.
Thinking ahead to a time where archival nodes are rare and UTXO only
nodes are common I wonder if there's a case for the block size limit to
be redefined in terms of the block's net effect on the size of the
UTXO. That is, there is no actual hard limit on the size of the block
itself. Only on the growth in UTXO that results from the transactions
Growth effect on the UTXO for a TX is approximately proportional to
number of outputs - number of inputs. So a consolidating transaction
that spends many inputs to few outputs may actually have a negative cost
to the UTXO size. The opposite being a UTXO growing transaction that
creates more outputs than it has inputs.
This creates an incentive for miners to include consolidating
transactions (which are good for the overall health of the UTXO) for no
fee because including them creates more space for growing transactions
which do come with fees.
I haven't thought it through in any detail. I'm sure there's a whole
range of issues to be considered around it. But short of 'the miner gets
to save some disk space' I can't think of any other economic incentives
for miners to offer a free 'cleanup service'.
Since we are already talking about things that are in the realm of a
hard fork. An alternative would be simply reserve a fixed percentage of
the block space (for arguments sake say 10-20%) that is only allowed to
contain consolidating transactions. Since this class of transactions
only has to compete with other consolidating transactions for block
space likely the fee/byte would be a lot lower and incentivize both
miners and users to keep the UTXO set as tidy as possible.
More information about the bitcoin-ml