I've tried to look it up when I was thinking about this and nothing came up. Mind sharing a link to a place that discusses it (as a soft-fork for Bitcoin)?


The while word there is important. This isn't trying to solve a problem of people losing their funds, but something different. Unfortunately, in this proposal adds a possibility of losing funds. The point is that erosion is very slow and starts only after a long time, so it should be trivial to avoid any loss whatsoever, and losing very little if mistakes happen (this is assuming you have access to you own prvkey at least once in a while, which now I understand is not valid in all cases).


I don't consider this telling people what to do with their money, since sending money between addresses in the same wallet is just a technicality. Everyone is still keeping their own money. Consider this: when you spend money from an address, Bitcoin protocol forces you to spend all the money from that address (keeping change by adding another output). Seems similar to me, in the sense that technically you're forced to do something particular with your money, but not really, since it's just details of the protocal that don't affect you (again, similar assuming you can access your prvkeys)

----

To summarize, seems to me at least some of you agree that the upsides of this proposal are worthwhile, but the trade-off (needing to move funds between different addresses in your wallet once in a while) is unacceptable, no matter what delay is chosen before erosion begins, or how slowly it occurs. I did not consider the scenario of people with cold wallets without access to prvkeys for very long durations, which is indeed a problem. I would never put myself in such a situation anyway, but others are free to do so. Otherwise, it's just a matter that people feel this requirement encroaches on their freedom of using their coins. I don't feel this way, but I can't quite argue with that either.

I agree, though this is trivial to solve. Simply exempt that one address from this rule. So there will be 21m minus whatever is burnt. Still has the advantages I listed.


Equating this idea with what you’re referring to (some altcoin implementing demurrage) only makes some sense if you think one transaction once every few years is expensive, which it isn’t, especially when you don’t care for quick confirmation. So if you do move your funds, you incur virtually no penalty (yes you do need to pay regular transaction fees though, I don’t think there’s a way to get around that - maybe they could be lowered by replacing regular transactions with some kind of signaling that is lighter in size, though that would probably complicate implimination). Once again, you're trying to say I'm suggesting people should lose funds over time, which completely misses the point. The whole point is to make it as easy as possible to not lose any money, while still being able to reusing lost funds (although at a significant delay unfortunately).

Also, has bitcoin not been laughed at time and time again? Don't argue through intimidation please.


You’re still making the same mistake. You know your public key has multiple different addresses, right? You are NOT forced to change your public key, private key, wallet, identity, number of children (that was kind of funny though), anything except the address at which the funds reside - you just send your funds from the address with the funds to another (fresh) address you own in the same wallet.


Well, OK, something to consider. I don't think personal temptations should enter the equation, but I guess this is a subjective matter.


What is mining? What are fees? You’re objecting to what Bitcoin already is. I’m not hoping they will trickle down anything. They obviously have to buy stuff with the bitcoins they mine (electricity, equipment, paychecks). Again, this is just what mining already is.

It starts to look like you're just intentionally misunderstanding me. I said multiple time that you would send to yourself. You are in no way forced to send coins to addresses of different wallets. I myself don't mind doing that at all, otherwise I wouldn't have proposed this in the first place.

Banks? Trickle down economic? Redistributing to the needy?
What are you talking about? Who is saying any of this?

---------

As for the legitimate argument made by others.


I admit I haven't thought about people creating multiple cold storage long term wallets for others to use in the far future that they cannot reasonably access every once in a while. However, know this is somewhat unsafe, as cryptographic methods have a tendency to break over time, by vulnerabilities being found and technological advances. It is likely you will need to migrate your wallet to eliminate a threat of theft, so keeping funds like that may not be a good idea. However, you should be free to make this decision... So ultimately I agree.

Thanks everyone except one of you for responding so politely *ehm ehm*
Here are my responses.


Interesting, learned something new.


You actually need a supermajority to activate this as a soft-fork, but yes I agree.


Well only if they can't bring themselves to make a simple transaction at the necessary intervals, but as Franky1 said there are situations where this is actually the case. So in those specific cases this is a problem, I agree.


Good point. What if locked coins don’t erode, i.e. count starts from the point the lock is released? This half-solves the point about the retirement funds / hodling, as you can lock funds and have them not erode. This also makes the proposal less effective, but not moot.


Once again, it only erodes if you fail to do a very simple procedure every few years. And if it helps stabilize the market price and incentivizes mining, then the pros could outweigh the cons, thus encouraging investment. As it is now, Bitcoin suffers from the same problem fiat and gold do, which is no one really knows how much exists of them.


They won't vanish after a few year, erosion should be very slow. The numbers I gave in the example are perhaps too high though, as I've mentioned whatever passes 95% consensus is what will eventually be implemented.


I don't see this as losing control in any way. It's just a stronger requirement of proof of possession.


The whole point of this proposal is to create such a way.


Yeah I also think the units thing is dumb, this isn't about that. This is about knowing how many units there are in circulation, to add transparency in the market.


I don't see it as using, since you're sending to yourself. It's more like signaling "hey, someone is still using these."


You would need a new field in the block to list the erosion effect, and a new kind of transaction (like Segwit) that can spend these eroded coins, both of these in places that are ignored by old client. Also new clients will not accept transactions that disregard erosion.
However, as I said I may be completely wrong here, and I am definitely not advocating a hard-fork for this.


Again, not about the units. The problem is no one knows if its 5% or 10% or 15% or 50%, so you estimate, which could blow up in your face one day. Imagine Satoshi starts spending their coins today...

Yes I agree, that is the downside of this suggestion. However, this seems acceptable to me. How much effort is it really, to broadcast a simple transaction every 2 year? It could alternatively be 5 year, or 10, and with slower erosion factor, whatever seems acceptable to 95% of users. Any erosion will increase transparency and security, it will just take longer to start and be less effective. Still an improvement from what we have now.


Yes maybe I should have been clearer about this here: the count will only begin after the soft-fork takes effect, so all users will have a good few years to hear about this change in consensus (I did write it in the link I posted btw). Haven't thought about people that have already intentionally locked their private keys away for many years without them being able to access them at all. It that a thing?

Just want to clarify once again, the point is not that people should hand out money after some time, only to redistribute funds that are actually lost.

I have a suggestion, and I'd like the community's thoughts.

It is well known that many bitcoins are no longer accessible due to loss of private keys, but no one knows how much. While estimates exist, they rely on imprecise methods. Most famously, there is still a large amount of bitcoin in Satoshi's wallet, and it is unclear whether they will ever be spent. This phenomenon causes undesirable uncertainty in the market, undercuts the slogan "there will eventually be ~21 million bitcoins", and in a sense causes deflation.

I thought of a way to increase transparency as to how many bitcoins are actually in circulation, and also increase miners' rewards, through a process that somewhat resembles erosion (hence the name). Seems to me it can be implemented as a soft-fork, though I haven't worked out the details to verify this (I'd like to hear what the community thinks about this idea before delving in more deeply, if ever).

Onto the idea itself. I'll make up some parameters as an example. Basically, from the moment an address receives funds, a countdown begins. Whatever funds are left in the address after 3 year will start eroding, meaning that every week from then on another 0.1% of the amount in that address may be taken by the miner (technically each block will contain a list of addresses with the funds that were taken).

If you have funds sitting in an address and you don't want them to erode, all you have to do is spend them to yourself once every 3 years. Hot wallets can automate this, and cold wallets can warn you to do this manually. Even if you forget, assuming you at least check once in a while that your funds are in place, you would see the erosion that began, and take action before much damage took place.

Although I don't think it's technically possible to include these eroded funds in the coinbase transaction itself (as that would make it a hard-fork), they could be implemented to act as if they were, i.e. they must mature 100 blocks before being spendable.

So to reiterate, we gain (a) transparency as to how many bitcoins are in circulation, since users are incentivized to prove that they still possess their private key every once in a while, and (b) miners will be able to access lost funds, returning them into circulation and increasing the security of the network.

I go into a bit more detail here to those interested: https://gist.github.com/yotamDvir/e7ff52a34460f3c9a6bf3051b0e51414.