[Bitcoin-ml] Block size credit and supporting "burst" block sizes
Steve
shadders.del at gmail.com
Mon Nov 20 08:54:37 UTC 2017
Agree with you Peter. I think it's an interesting thought experiment to
imagine what would happen to fee dynamics if block size limit was either
removed completely or raised to some very high value that we aren't
going to forseeably hit in the next few years (e.g. 1GB).
Miner enforced soft limits are a mechanism we haven't seen explored in
practice yet. Yet they could function almost exactly like the existing
hard limit. The only difference being that it is the miners that decide.
It is a mistake to equate this scenario with unbounded supply and assume
fees would crash. Mining a transaction has a quantifiable cost of
inputs. If it is unprofitable to mine a transaction a rational miner
will ignore it. So there is a natural floor price already built into
the system. Of course until block rewards become a minimal proportion of
revenue this effect won't be very visible. Similarly there is a natural
ceiling price also built into the system by competition. If a group of
miners insist on constraining block size to push up fees it only takes
one miner to break ranks and mine larger blocks or lower fee
transactions then the others have little choice but to follow suit or
lose market share. This only works with an effectively unbounded block
size limit. A rogue miner with say 5% of the hashrate may only mine a
block every 3 hours but it can be twenty times the usual size and they
can clear the entire mempool when they do. The incentive system exists
for miners to push transactions fees up up to the point where those fees
start to become profitable but the ability to continue doing so tapers
off rapidly after that. I envisage mining will become an extremely
competitive business with thin margins.
I think with a very large or unlimited block size limit we would see
this bursting behaviour emerge naturally. In theory we are already
protected from spam by the miners. They are free to reject any
transactions that are unprofitable to mine. If they are profitable who
cares if it's spam. In practice though the technical challenge of
customizing transaction selection policies means that miners are not
being as responsive with transaction selection as they could be. One of
the things on my (already too long) TODO list to make transaction
selection a plugin type process so that miners can use an API and their
language of choice to write their own transaction selection algorithm
and get a bit more sophisticated with this side of mining.
On 20/11/2017 3:34 PM, Peter Rizun via bitcoin-ml wrote:
> (That said, if miners or node operators want to create a sort of soft limit within these constraints, through some decentralized signalling mechanism, a la BU’s EC or something else, I would see that as a beautiful example of emergence. The point is that the block size limit should not be seen as a policy tool, and certainly not as a tool to be wielded by the developers.)
>
> Peter
>
> _______________________________________________
> bitcoin-ml mailing list
> bitcoin-ml at lists.linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-ml
More information about the bitcoin-ml
mailing list