But now (or Soon) is not the right time. We need studies, simulations, and (most importantly) actual empirical feedback from persistent full blocks to best determine how and when to proceed with altering the 1mb parameter.
Wrong. We already know it will be a
clusterfuck.I venture my opinion from 30 years experience in IT, when it appears that you have zilch and should just speak to what you know about (Monero? Hashfast?)
good re-read. and i'm pretty sure he changed that write-up. initially, he claimed the entire UTXO was held in RAM but down in the Reddit comments for the thread several ppl pointed out that it was held on disk with a 100MB high speed cache. so, bottom line, it doesn't necessarily appear that this is a problem except for maybe miners. given that tx growth won't immediately go to 20MB/block, i think it's safe to say this space problem should be worked out in time.
I really think there is a fast and simple constraint on UTXO bloat which can be done:
Allowing the existing free transaction space to be used for tx which reduce UTXOs (i.e. negative delta) instead of being based upon the number of days destroyed, which was to encourage old coins being spent, something less important.
