Quote from: zimmah on October 16, 2016, 08:09:34 AM

they would kill bitcoin since they steal mining fees and pocket it for their investors. This leaves miners without fees, and it would kill bitcoin in the long run, once block rewards become too small to cover the expenses for miners.

right now miners are not reliant on tx fee's. the miners income is over 97.6% reward:2.4% fee (12.5btc:0.3btc) (reward:txfee) it will be a long time before the flip occurs where miners are reliant on tx fee's and by that time the 'popularity' of bitcoin should have improved enough to have lots of opening and closing of channels to pay the miner fee's.
todays 0.3 fee total becomes 50:50 after the year 2024 (2 reward halvings) but before 2028
changes in bitcoin valuation and also blockcapacity also change/delay when the flip will occur.

lets say we are in the future where 1btc is $6000 and say miners want $10k in just fee's (as oppose to ~$7,500 in reward now) per block..
they will require blocks capable of 166,666tx ONCHAIN at ~6cents a tx(<70mb block)acceptable bloat in 20 years, acceptable fee
they will require blocks capable of 8,500tx ONCHAIN at ~$1.20cents a tx(<4mb block)acceptable bloat in 8 years, not acceptable fee

im not going to bother calculating the 70mb block scenario of 20 years.because although the spread of fee's is cheaper its just going to make the blockstreamers knitpick the blocksize debate. rather than the point about fee's

now lets deal with the finer details
its worth noting
the average person uses a debit/credit card only 20 times a fortnight in the real world, which breaks down as 12cents a LN tx if the same usage occurred offchain within LN to cover the combined $2.40 open/close transactions per fortnight
so the presumption is for LN to remain open for longer to reduce overall cost of use per offchain tx.

which if people were doing the 20tx a fortnight. means that based on a 4mb block in 8 years time
UPTO 2125 channels open per block(2peers tx pay in=1channel), 4250 channels close per block(both paid out in one tx)
a fortnights lock of 4250 people. which equates to a POSSIBLE 8 million people using LN MAX (at 12cent/tx for their usual 20tx a fortnight).
a months lock of 4250 people. which equates to a POSSIBLE 16 million people using LN MAX (at 6cent/tx for their usual 40tx a month).

ofcourse, of block capacity was 8mb in 8 years the onchain fee per tx would be $0.60 to give miners $10k (allowing upto 32mill users using LN)
ofcourse, of block capacity was 16mb in 8 years the onchain fee per tx would be $0.30 to give miners $10k (allowing upto 64mill users using LN)

summary:
i see it as possible to work but thats assuming that
8-16mill active bitcoiners exist
8-16mill people will be happily even want to lockfunds in.
8-16mill people will be happily do 20tx a fortnight-month usage to equate to 6-12cents each LN tx, to cover the combined $2.40 open/close tx
miners want only $10k - if they want more fee's get higher and timelocks need extending to keep overall cost of use down
bitcoin valuation is $6000 - if its lower fee's get higher and timelocks need extending to keep overall cost of use down

and now lets wait for the mindless, unbacked up, unresearched blockstream defender rants that blockstream desire to cause fee war in 2017 instead of 2024-2036 so that offchain/sidechain are the permanent peg utopian future and eventually no one should use bitcoin onchain transactions because they see large fee's are a good barrier of entry to use bitcoin and a ploy to get people to go offchain. (usual bitcoin killing mantra)