Sorry, if a raw dump doesn't serve the context exactly, but this is sort of related:
Hey all,
there was some discussion on GitHub related to OP_RETURN and transaction
costs. One point that was raised was the cost associated with output
amounts. While what has been said holds true, there are actually a few
options to save costs.
I'd like to share, and try.., a new "format":
https://i.imgur.com/hiZFZa7.png (overview)
Chancecoin demonstrated with several thousand transactions that both
output amounts as well as transaction costs can be cut significantly
without adopting a new encoding scheme, but simply by reducing output
amounts and fees.
I tracked over 2'100 transactions from broadcast to confirmation and
mapped most blocks to mining pools. It turns out that transactions with
output amounts as low as 0.00000780 BTC and transaction fees of 0.00001
BTC confirm within about 20 minutes on average while 85 %+ pools (in
terms of hash power/total network) are mining those transactions. This
includes GHash as well as DiscusFish.
The actual data can be found here:
https://docs.google.com/spreadsheets/d/1ELqeBlKKsze6Tya-Arh6HfoFcnV5CAtAObmhEb_wmSM/ TL;TR + conclusion:
Given that at least 75 % of all nodes are using Bitcoin Core v0.9+ and
about 85 % pools confirm ultra cheap transactions in general, I think
it's time to step back from the old "0.00006 BTC/packet rule" and adopt
output amounts of 0.00000546-0.00000882 BTC which are defacto default
and standard since Bitcoin core v0.9.
Furthermore it might be worth a try to cut down transaction fees from
0.0001 BTC/1000 byte to 0.00001 BTC/1000 byte transaciton size, given
the reasonable confirmation delay of about 20 minutes. The later should
not be enforced, but could be offered as an experimental option for
users who may choose cost savings over confirmation speed.
Cheers!
