I'm now *completely* convinced I'm incapable of having an original idea that hasn't already been posted on the forums. I arrived at a similar idea trying to deal with the issue of lost coins and blockchain pruning. Presumably if they're valuable enough, people hopefully won't lose them as much, but, stuff happens. I am assuming a future where bitcoin is mainstream and widely used and distributed. An unknown trickle of coins will vanish into the bitbucket over time, forever unavailable *and* forever occupying space in the blockchain and UTXO. This will be cruft that must be carried forward 10, 20, 50 years? Here's my thoughts, assuming this idea hasn't been advanced in some thread I haven't managed to read.
What if a form of BTC expiration were built into the protocol? Say unspent coins have a lifetime of 8 years, or, more specifically, 420000 blocks (twice the mining reward change cycle). If not moved during that time, their value expires and goes to the miner of the next block. This would have several effects:
1) putting truly dead coins back into circulation, so that all of the 21M coins will be available into the indefinite future.
2) guaranteeing the maximum blockchain length will be 420k blocks. Apart from archival history, it would be impossible to have a spendable transaction older than that.
3) In the event tx fees can't fully support mining efforts as the block rewards shrink, it will be an equitable additional source of miner income.
There's plenty of time to get the word out (better sooner than later, before bitcoin really takes off). There's plenty of time to sort out implementation details. While I hate to make an analogy with the traditional banking system, it's similar to unclaimed account laws where idle accounts eventually go to the state. Of course the biggest issue is that it changes the terms on all existing bitcoin owners in an unexpected manner. It would suck for the guy who in 2010 put some BTC in some time capsule that won't be opened until after 2020. It would be problematic for physical bitcoin like Casascius. The alternative is that in 2060, we'll have a multi-terabyte blockchain that would need to be fully rescanned to prove the output from block 1 is spendable...
what if you make your own coin with such protocol changes.. and see if your own coin takes off.. call it byte coin with the moto something like "its like bitcoin, but 8 years in length". (EG bit x8=byte)
no one want to hear their bitcoins would evaporate within 8 years. by having
bitcoins protocol that can delete TX's that are 8 years old and then increase the standard block reward to cover the amount deleted would just make it so bitcoins can be reprinted.. which would ruin bitcoin
so how about make your own coin and see if it catches on, not change bitcoins protocol
scenariooh and by the way, that $20k FIAT your grandma has put away safely under her matress for the last decade. its now gone.. tell her its been burnt, she cant have it back, but now someone else is $20k richer just by being the lucky worker at the US mint that day the certain printing batch came out that had an extra $20k in it. tell her you allowed FIAT law to change so that if she didnt spend it in 8 years she would lose it, and its all your fault.
or better yet. tell her theres a new currency coming out that has a 8 year lifespan, she can put her money under the matress into. see if she would buy into it or stick with the old unchanged rules that means what is her's is hers for life
as for worrying about blockchain sizes..
in 1999 a 4gigabyte hard drive was considered big. a decade later a terrabyte hard drive was considered big. 5 years after that 2terrabyte hard drves can be bought for under $100.
now then at 1mb block. if every block was filled to the top. that would be 52gb a year growth of the blockchain*. so far in 4 years we have not even gone that far. and lets say in 10 years it did manage to contain 100% full blocks that is only half a terrabyte.
im sure in the next 10 years the price of terrabyte hard drives and internet speeds would make the potential blockchain length a non-issue. so again bitcoin protocol changes are irrational and would cause more damage then help.
*6mb per hour x24 = 144mb per day x 365 = 52.56GB per year