The premise that fees will replace subsidy without fundamentally threatening the network's security is still untested. It's theoretically possible that waning future block rewards could lead to hash rate drops that throw Bitcoin's byzantine fault tolerance in question. Severe enough drops in hash rate could allow older generations of hardware to come back online, opening up the possibility of massive block reorganizations -- and therefore utter unreliability as a value/monetary system.
Massive block reorgs would most likely happen if someone decides to execute an attack on Bitcoin. Old and obsolete ASICs are still able to come back as of now. There isn't any relation to older ASICs coming online, as far as the security of the network is concerned. If you're going to attack the network, might as well as use the most efficient ASICs out there, no need for old ASICs.
Maybe you missed the point about "hash rate drops that throw Bitcoin's byzantine fault tolerance in question." I wasn't making a statement about potential attacks on the system
as is. The context is dwindling block reward (due to lack of fee income), which disincentivizes mining and sends hash rate into a downward spiral. Consider
this IRC discussion from 2015:
gmaxwell: Lets imagine the most 'sacred' parameter, the supply of coins. Now imagine a future 25 years from now where subsidy is very low, and TX fees are not picking up the slack (e.g. fee market has failed or is insufficient) and the security of the system is failing, the network is being reorged by byzantine attackers with generation old hashpower. Security and usability are evaporating. Something must be done. With bitcoin's utility failing, saying you now need to pay fees when you haven't the last 10 years may not be a credible argument. So what do you do?
I'm really bad at economics, so forgive me if I misunderstood anything. Isn't the additional coin going to affect the market? If its too small, it wouldn't make a difference to the miner.
Of course it will affect the market. So does Bitcoin's current and historically high inflation rate. The point was that it
might have made economic sense -- particularly from a long term perspective -- to pay (via inflation) for better security guarantees. Those security guarantees may in turn have made your investment more valuable in the long run. I assume this is why
Peter Todd feels this way:
IMO Bitcoin should have had an explicit 1%/year or so security tax, implemented via inflation... [1/2]
The main thing that attract most of us is that Bitcoin has a fixed and transparent coin supply. I doubt most people would support breaking the basic feature that defines Bitcoin.
Yes. Unless the protocol became fundamentally broken, in which case, users might be incentivized to fix it. We'll see how Bitcoin fares when the subsidy begins to run dry:
It's theoretically possible that waning future block rewards could lead to hash rate drops that throw Bitcoin's byzantine fault tolerance in question. Severe enough drops in hash rate could allow older generations of hardware to come back online, opening up the possibility of massive block reorganizations -- and therefore utter unreliability as a value/monetary system.