A proposed solution for lost Bitcoins...

[odolvlobo]

Regarding the "reclaiming blockchain space" problem, there was a research project some years ago, the "Mini-Blockchain project", which would have allowed Bitcoin to run only with pruned nodes. The proposed solution, however, changes the transaction format into one more similar to Ethereum's (account-based instead of UTXO-based) and doesn't allow complex contracts with Bitcoin Script. It is implemented in some altcoins.

The "UTXO commitment" idea comes into my mind, also. In this solution, the UTXO set is hashed and commited to every block, allowing nodes to go online without having to validate the rest of the blockchain. In theory, this could allow mini-blockchains, too - but I think the challenge would be how to handle orphans/reorgs, so at least some hundreds or thousands of blocks would have to be kept.

For other purposes than the blockchain size issue, reclaiming old coins makes no sense for me, like for the other participants in the discussion.

My idea is different from prune, i try to only delete genesis block always that i can and estabilish a always a new genesis block.
At same time the chance of could recover address with lost bitcoins and more then 100 years not cashed out.

To really solve full-node really space problems i think maybe this can be better: https://bitcointalk.org/index.php?topic=5070680

What happens to the coinbase transaction in a block that you are going to delete?

[cfbtcman]

What happens to the coinbase transaction in a block that you are going to delete?

If there is no BTC yet in the source and target addresses there is some problem deleting all the block?

[d5000]

My idea is different from prune, i try to only delete genesis block always that i can and estabilish a always a new genesis block.
At same time the chance of could recover address with lost bitcoins and more then 100 years not cashed out.

To really solve full-node really space problems i think maybe this can be better: https://bitcointalk.org/index.php?topic=5070680

If you're trying to find a "scaling solution", you first have to think about the exact problem you want to solve.

Your solution to simply "delete old blocks" solves primarily the "IBD" (Initial Blockchain Download) problem, i.e. the problem that every full node has to sync to the full blockchain, which costs lots of bandwidth (but isn't a problem at all for nodes that are continuously connected to the network).

That's why I mentioned the Mini-Blockchain project (implemented in a small altcoin called Cryptonite) and UTXO commitments - they solve the exact same issue, but without having to delete transactions (which some see, not without reason, as "stealing"). However, they have other kinds of drawbacks, as I intended to explain.

Your solution also would collide with other scaling proposals, for example, LN and sidechains. For example, a long-running, stable LN channel would run into danger once its channel-opening transaction is near the "deadline". LN implementations would have to re-open the channel in this case, which makes the apps overly complex. The same applies to sidechains: old "burning transactions" that result in "sidechain coins" being created (search for 2-way pegs for an explanation, e.g. here) would be at risk to be "forgotten", too. You could establish special rules for these cases - but again, this results in higher complexity.

However, I encourage you to simply try out the concept, if you're into coding. Fork the Bitcoin code and implement it - that's what altcoins are for, in my opinion If you coin is accepted on the market, then it may become an interesting addition for Bitcoin, too. However, I see only drawbacks with respect to the already proposed solutions (not only LN).

Maybe there is need for a "Scaling solutions" thread, listing LN, sidechains, sharding et al. - so people know what proposals are and the drawbacks of every proposal. In ancient times I created a thread about cryptocurrencies that intend to solve the "scaling" problem, but the scope is slightly different - e.g. it does not cover solutions that are still only concepts and the whole bunch of "2nd layer" proposals from the Bitcoin community.

[cfbtcman]

Your solution to simply "delete old blocks" solves primarily the "IBD" (Initial Blockchain Download) problem, i.e. the problem that every full node has to sync to the full blockchain, which costs lots of bandwidth (but isn't a problem at all for nodes that are continuously connected to the network).

That's why I mentioned the Mini-Blockchain project (implemented in a small altcoin called Cryptonite) and UTXO commitments - they solve the exact same issue, but without having to delete transactions (which some see, not without reason, as "stealing"). However, they have other kinds of drawbacks, as I intended to explain.

Your solution also would collide with other scaling proposals, for example, LN and sidechains. For example, a long-running, stable LN channel would run into danger once its channel-opening transaction is near the "deadline". LN implementations would have to re-open the channel in this case, which makes the apps overly complex. The same applies to sidechains: old "burning transactions" that result in "sidechain coins" being created (search for 2-way pegs for an explanation, e.g. here) would be at risk to be "forgotten", too. You could establish special rules for these cases - but again, this results in higher complexity.

Pruned nodes should work other way, my proposal only deletes genesis blocks without BTC inside all addresses included in the transactions of the block and  even now in actual genesis block there is BTC inside, for example this address:

1A1zP1eP5QGefi2DMPTfTL5SLmv7DivfNa

Satoshis genesis address haves money inside, if the guy was really Hal Finney and he is dead this will be lost bitcoins forever, so, in 100 years my proposal would delete genesis block.

That would only afect one part of the balance of that address, it will not stay empty, only without the first 50 bitcoins and that 50 bitcoins would go to miners rewards.

Satoshi said that after all 21 millions was produced the miners would earn the transactions commissions, what transaction comissions if now the only solution for bitcoin scaling is the LN ?

So, all the money that stays in the network after people dies and left the bitcoins forever, should be inserted again in circulation to power new miners and nodes, why i should keep mining or keep a bitcoin core node if everyone is using 2nd layer?

Maybe even the best solution is putting less than 100 years, like the people do in the banks, after 7 years the money goes back to the system if the guy dies and nobody could reclaim it.

And even better than everything, i thought about that now, WE WOULD KNOW IF SATOSHI IS STILL ALIVE AND LOOKING OR HE IS REALLY DEAD!

[odolvlobo]

What happens to the coinbase transaction in a block that you are going to delete?

If there is no BTC yet in the source and target addresses there is some problem deleting all the block?

I guess what I don't understand yet is how do you validate a transaction who's inputs come from a deleted block? I guess maybe you just assume it is valid because the transaction is so old. That's an interesting concept. Normally, you base validity on the genesis block, but in this case you assume validity based on the current block.

There is also an issue of determining the longest chain, since every chain will have the same length.

[mocacinno]

I guess what I don't understand yet is how do you validate a transaction who's inputs come from a deleted block? I guess maybe you just assume it is valid because the transaction is so old. That's an interesting concept. Normally, you base validity on the genesis block, but in this case you assume validity based on the current block.

This ^^ is the point i was trying to make... As soon as you throw away blocks or transactions, you'll have to start using a trust based validation system... In my personal opinion, this isn't a good basis for validation. It doesn't matter if a group of 1.000.000 trusted individuals provide a trustbase for everybody else, it's still not good enough for me... I want to be able to mathematically verify the complete chain right from the genesis block by myself. As soon as the information is removed, there is no way for me to do this verification and in my honest opinion, at this point the system becomes broken.

There is also an issue of determining the longest chain, since every chain will have the same length.

Interesting point, i didn't even think about this yet, but this observation is also a valid counter-argument.

[cfbtcman]

What happens to the coinbase transaction in a block that you are going to delete?

If there is no BTC yet in the source and target addresses there is some problem deleting all the block?

I guess what I don't understand yet is how do you validate a transaction who's inputs come from a deleted block? I guess maybe you just assume it is valid because the transaction is so old. That's an interesting concept. Normally, you base validity on the genesis block, but in this case you assume validity based on the current block.

There is also an issue of determining the longest chain, since every chain will have the same length.

How you validate the transactions from the genesis block? From what block comes the input?

About longest chain, the block database dont have one ID number or dont use date?

If i put one new genesis block with reverse engineer i would fake all the network?

[d5000]

Satoshi said that after all 21 millions was produced the miners would earn the transactions commissions, what transaction comissions if now the only solution for bitcoin scaling is the LN ?

LN (and sidechains, too) cannot "live" without on-chain transactions - LN needs them to open/close channels, and sidechains to burn/unburn coins transacted between different sidechains.

Also, on-chain transactions have always advantages versus second-layer ones: they're totally trustless, no attacks like "channel rewinding" or "weak sidechain becomes attacked by miners" are possible. The main disadvantage, the confirmation time, is irrelevant for large transactions.

In a scenario where really "everybody is using LN", on-chain transactions could become very cheap and no transaction fee would be needed. As on-chain tx do have advantages over off-chain tx, there will be people using them (apart from channel-opening/closing transactions).

In short: if the on-chain transactions are cheap, blocks will fill again fast, and transaction fees will rise. So the transaction fee should self-regulate and be greater than zero. In a massively used Bitcoin environment miners will still have significant income.

So, all the money that stays in the network after people dies and left the bitcoins forever, should be inserted again in circulation to power new miners and nodes, why i should keep mining or keep a bitcoin core node if everyone is using 2nd layer?

I understand your point, and maybe with a very long timeframe something like this could work (>10 years). It's not a bad idea, but I think it won't be implemented in a long-running coin like Bitcoin. Try to build an altcoin with that proposal and look if it's being accepted.

There are some concepts that have some similarities: In Freicoin, "hoarded" coins must pay a demurrage fee. I don't know the exact rate but Satoshi's coins would be worth much less today if Bitcoin had implemented such a system. These coins could be returned to a fund and be paid out to miners.

The problem of Bitcoin's "lost" coins is that nobody knows that they're really lost - they're parked in a sort of limbo, as they could be reactivated at any time. So we don't know if they form part of the available supply or not, and the problem will aggraviate with time. This is, for me, the strongest argument for your proposal.

I guess what I don't understand yet is how do you validate a transaction who's inputs come from a deleted block? I guess maybe you just assume it is valid because the transaction is so old. That's an interesting concept. Normally, you base validity on the genesis block, but in this case you assume validity based on the current block.

Many proof-of-stake coins do something very similar - they use a checkpoint system where reorgs longer than X are simply not allowed. If X=1000, then the block 1000 blocks before the current block becomes the earliest block from which a "tree" of different chains is allowed to exist, and all blocks before are considered "finalized". Chains that depart from the tree earlier are ignored by clients. So transactions that are confirmed in blocks before this last "finalized" block are assumed to be valid (they're also finalized).

The same thing would occur in cfcbitman's system. However, I think an account-based system like Ethereum's (instead of an UTXO-based one like in Bitcoin) would be advantageous.

There is also an issue of determining the longest chain, since every chain will have the same length.

I think not - afaik the chain length would be determined by the accumulated difficulty since the last block which is considered "finalized". (Maybe I understand your issue wrong)

[cfbtcman]

I know 0.00000000000000000001 its possible, but its not pratical and human, im not trying to go back to traditional banking system and i agree with you that USDT its a scam, but if we could turn bitcoin payments more human stopping using a lot of decimal parts why not?

If in the end of the 21 millions we make one airdrop to double the bitcoins of everyone or multiply by 100000000 that dont makes any difference and the bitcoin would look more easily to normal people figure out is value, if you could buy almost exactly the same things with one bitcoin and 1 dolar you will have a better idea of the money you are wasting without doing calculations 0.00001 x $5700 = ?

As mentioned above:

Hence why you got mBTC, uBTC, Satoshis, etc. The question of terminology of fractions of Bitcoin is long solved and a non-issue.

The usage of mBTC (ie. 0.001 BTC) is already fairly common. I think some wallets even default to displaying your balance in mBTC rather than BTC, precisely to get rid of those decimals

Currently mBTC 1.00,- equals USD 5.43,- which is not much different from calculating foreign fiat currency exchange rates. For smaller amounts you can use uBTC (sometimes called bits) with uBTC 1.00,- currently equaling about half a penny. Then there's sat (satoshis), msat (millisatoshis)... and just like that you get ever smaller, manageable units of account.

Too much units to use and change, more confusing, not human, with bitcoin flutuation nobody can use that units pratically, i know wallets that use it and i immediatly change to Euro/Dolar or BTC units, much better.

If Fiat money finished and bitcoin was the only coin in the world, somebody would create a FIAT money like dolar to be adopted as currency unit and would have great sucess.

If BTC wants to be solution for the world as the only coin, then needs to create more than 21 millions units and make the value +- equal to dolar to the people start to wean of dolar/euro.
Euro coin was one example, they created it with the same value as dolar, there was a reason, USDT the same, etc.

That could be done very easily, just multiply the addresses that have money-in by some number.

[mocacinno]

--snip--
Too much units to use and change, more confusing, not human, with bitcoin flutuation nobody can use that units pratically, i know wallets that use it and i immediatly change to Euro/Dolar or BTC units, much better.

If Fiat money finished and bitcoin was the only coin in the world, somebody would create a FIAT money like dolar to be adopted as currency unit and would have great sucess.

If BTC wants to be solution for the world as the only coin then needs to create more than 21 millions units and make the value +- equal to dolar to the people start to wean of dolar/euro.
Euro coin was one example, they created it with the same value as dolar, there was a reason, USDT the same, etc.

That could be done very easily, just multiply the addresses that have money in by some number.

As mentioned above, i personally feel this is a non-issue aswell...
If you're really bothered with the current units, fork one of the open source wallets, edit the sourcecode by multiplying the number that's shown when the wallet show BTC by 100000000 and change the name to satoshi...

Most wallets even use mBTC µBTC and satoshi as potential units... Nothing is stopping you from editing the sourcecode and removing mBTC and µBTC (only leaving satoshi as potential units).

For all i care, fork bitcoin core, work with mSatoshi and call 1 mSatosi 1 cfbtcman... This way you'll be able to send 10.000.000 cfbtcman's to somebody else... Ofcourse, this other person will just see an incoming transaction funding his address with 10.000 satoshi ... Who knows, maybe your wallet will become popular...

Last but not least, the floating point values are a non issue as well, every time i develop anything crypto related, i always count in satoshi (an integer, worth 1/100000000 of a BTC). There's nothing stopping you from doing the same.

[pawanjain]

I've discussed this in the past, but my point of view in this matter is very simple: you can't drop any blocks...
As soon as the complete network dropped the first 100 blocks (for example) nobody would ever be able to re-verify the complete blockchain, so you'd have to start trusting other people instead of being able to verify everything yourself...

You wouldn't have to trust other people, if for example the first 500000 blocks would be verified by enough nodes and a SHA512 checksum (or some other way to make a cheksum) of the verified 500000 blocks and the block data would be made available to download from outside the blockchain. Then those blocks can be deleted from the blockchain. All the new blockchain would need is the blockstate of the moment of block 500000 and the cheksum.
It will happen sooner or later. There is no way to avoid it.  

And there is no risk of permanently losing the data in 500000 first blocks, because every service tracking blockchain and origins of bitcoins would certainly want to keep a copy of that info. NSA is one of them.  

As for the OP's question: i agree with odolvlobo, it doesn't matter if people don't give their keys to their heirs... At the moment, the smallest unit is one satoshi (0.00000001 BTC), nothing is stopping the community from implementing even smaller units (for example, one  millisatoshi). The more people lose access to their funds (for example by forgetting or dying), the less BTC will be available, the higher the price will be...
It's perfectly possible to run a complete community whith only 1 BTC in circulation... Ofcourse you'll probably be paying in micro-satoshi instead of mBTC.

It wouldn't really work with just 1 BTC. The reason is that the whole bitcoin market value would flip if someone succeeded in recovering just 10 BTC from his old wallet.
We would suddenly have 10x the previous amount of bitcoins and the value would collapse. Even just the possibility that someone could find more bitcoins would be enough to scare investors away.

But yeah, if there was only 1 BTC and no risk of anyone finding more, then it would work.

I am trying to grab as much understanding from this post as I can. I do have a doubt here and that is whether the 'blockstate' of that particular block and it's corresponding checksum would be sufficient to verify the entire block ? I mean that would it be possible to know how much bitcoins were transacted to which addresses and at what time. If so then Yes this would definitely save the disk space.
Also, don't you think that keeping the block of data outside the blockchain would be risky? We do still have the option of keeping the pruned blocks on a different blockchain or a second blockchain.

[MarketNeutral]

How can one assume that the bitcoins are indeed lost? Don't some people keep valuables in bank vaults for many decades? Don't others bury valuables in their backyards? Gold bullion is passed down through generations without ever leaving the vault or seeing the light of day. Grandpa's old stash of cash that he hid in the crawlspace is still legal tender (at least in countries that didn't hardfork their currency). What if I want to pass down bitcoins to my children, but they're still young and cannot yet be entrusted with the private key? What if a private key is held in trust for decades? What if the bitcoins I mined in 2010 are still under my control, and I intend to keep them indefinitely? As regards the quantity in circulation, decimalization allows for nearly any quantity of bitcoins to circulate without issue. Why assume that coins are lost if they haven't moved for an arbitrary amount of time?

This is a solution looking for a problem.

[figmentofmyass]

How can one assume that the bitcoins are indeed lost? Don't some people keep valuables in bank vaults for many decades? Don't others bury valuables in their backyards? Gold bullion is passed down through generations without ever leaving the vault or seeing the light of day. Grandpa's old stash of cash that he hid in the crawlspace is still legal tender (at least in countries that didn't hardfork their currency). What if I want to pass down bitcoins to my children, but they're still young and cannot yet be entrusted with the private key? What if a private key is held in trust for decades? What if the bitcoins I mined in 2010 are still under my control, and I intend to keep them indefinitely? As regards the quantity in circulation, decimalization allows for nearly any quantity of bitcoins to circulate without issue. Why assume that coins are lost if they haven't moved for an arbitrary amount of time?

This is a solution looking for a problem.

i can see the value in it---what could be better for money than a completely predictable (but scarce) supply? it's just not something that could ever be implemented in bitcoin. too controversial. it's an interesting idea for an altcoin though. the protocol could periodically burn or redistribute unspent coins as miner fees, depending on the preferred outcome (encourage either hoarding or spending). this would have the added benefit of drastically reducing the size of the UTXO set.

aside from verifiable burn addresses, we obviously can't assume anything for sure. but people will continue looking for solutions to the problem of unpredictable supply. bitcoin's hard capped supply was a novel take on that. i suppose efforts like these try to take it to the next level.

[cfbtcman]

--snip--
Too much units to use and change, more confusing, not human, with bitcoin flutuation nobody can use that units pratically, i know wallets that use it and i immediatly change to Euro/Dolar or BTC units, much better.

If Fiat money finished and bitcoin was the only coin in the world, somebody would create a FIAT money like dolar to be adopted as currency unit and would have great sucess.

If BTC wants to be solution for the world as the only coin then needs to create more than 21 millions units and make the value +- equal to dolar to the people start to wean of dolar/euro.
Euro coin was one example, they created it with the same value as dolar, there was a reason, USDT the same, etc.

That could be done very easily, just multiply the addresses that have money in by some number.

As mentioned above, i personally feel this is a non-issue aswell...
If you're really bothered with the current units, fork one of the open source wallets, edit the sourcecode by multiplying the number that's shown when the wallet show BTC by 100000000 and change the name to satoshi...

Most wallets even use mBTC µBTC and satoshi as potential units... Nothing is stopping you from editing the sourcecode and removing mBTC and µBTC (only leaving satoshi as potential units).

For all i care, fork bitcoin core, work with mSatoshi and call 1 mSatosi 1 cfbtcman... This way you'll be able to send 10.000.000 cfbtcman's to somebody else... Ofcourse, this other person will just see an incoming transaction funding his address with 10.000 satoshi ... Who knows, maybe your wallet will become popular...

Last but not least, the floating point values are a non issue as well, every time i develop anything crypto related, i always count in satoshi (an integer, worth 1/100000000 of a BTC). There's nothing stopping you from doing the same.

Why should be me changing something and hardforking alone?
That would be the dead for bitcoin if any supporter that thinks something should be changed starts to make is own changes and hardforking, in 100 years we would have 1 million of coins, this needs to be presented and supported by bitcoin core people and we should stay united.

[cfbtcman]

I've discussed this in the past, but my point of view in this matter is very simple: you can't drop any blocks...
As soon as the complete network dropped the first 100 blocks (for example) nobody would ever be able to re-verify the complete blockchain, so you'd have to start trusting other people instead of being able to verify everything yourself...

You wouldn't have to trust other people, if for example the first 500000 blocks would be verified by enough nodes and a SHA512 checksum (or some other way to make a cheksum) of the verified 500000 blocks and the block data would be made available to download from outside the blockchain. Then those blocks can be deleted from the blockchain. All the new blockchain would need is the blockstate of the moment of block 500000 and the cheksum.
It will happen sooner or later. There is no way to avoid it.  

And there is no risk of permanently losing the data in 500000 first blocks, because every service tracking blockchain and origins of bitcoins would certainly want to keep a copy of that info. NSA is one of them.  

As for the OP's question: i agree with odolvlobo, it doesn't matter if people don't give their keys to their heirs... At the moment, the smallest unit is one satoshi (0.00000001 BTC), nothing is stopping the community from implementing even smaller units (for example, one  millisatoshi). The more people lose access to their funds (for example by forgetting or dying), the less BTC will be available, the higher the price will be...
It's perfectly possible to run a complete community whith only 1 BTC in circulation... Ofcourse you'll probably be paying in micro-satoshi instead of mBTC.

It wouldn't really work with just 1 BTC. The reason is that the whole bitcoin market value would flip if someone succeeded in recovering just 10 BTC from his old wallet.
We would suddenly have 10x the previous amount of bitcoins and the value would collapse. Even just the possibility that someone could find more bitcoins would be enough to scare investors away.

But yeah, if there was only 1 BTC and no risk of anyone finding more, then it would work.

I am trying to grab as much understanding from this post as I can. I do have a doubt here and that is whether the 'blockstate' of that particular block and it's corresponding checksum would be sufficient to verify the entire block ? I mean that would it be possible to know how much bitcoins were transacted to which addresses and at what time. If so then Yes this would definitely save the disk space.
Also, don't you think that keeping the block of data outside the blockchain would be risky? We do still have the option of keeping the pruned blocks on a different blockchain or a second blockchain.

pawanjain my man, now we start to talk "english"!
We need to start to cypher bitcoin balance in each block transactions and delete old blocks or force the users to move the money every X years and make a timeframe blockchain, every new day, we delete everyday blocks of X years ago.

Like Aplister said (Aplister you earned my respect, i need to met you, this is like Einstein-Eddington partnership loool) if somebody could recover 10 BTC someday after we use only one for all the world that would be a colapse.

[cfbtcman]

How can one assume that the bitcoins are indeed lost? Don't some people keep valuables in bank vaults for many decades? Don't others bury valuables in their backyards? Gold bullion is passed down through generations without ever leaving the vault or seeing the light of day. Grandpa's old stash of cash that he hid in the crawlspace is still legal tender (at least in countries that didn't hardfork their currency). What if I want to pass down bitcoins to my children, but they're still young and cannot yet be entrusted with the private key? What if a private key is held in trust for decades? What if the bitcoins I mined in 2010 are still under my control, and I intend to keep them indefinitely? As regards the quantity in circulation, decimalization allows for nearly any quantity of bitcoins to circulate without issue. Why assume that coins are lost if they haven't moved for an arbitrary amount of time?

This is a solution looking for a problem.

Actually some banks take your money out after few years without moving account.
Ok can exist some  account types that can be years without moving, but bitcoin can have is own rules.

If you know that you need to move the money every 10 years to say (im alive) and improve network operation and prevents bitcoin cataclism generated if someone recover a big BTC quantity address lost in time why we shouldnt do it?

Look what happens just in 10 years that everybody is fearing that the really Satoshi starts to sell all his bitcoins?
Look what Craig Wright said that he could put bitcoin in $1.000 if he wants and got many attention from the media and possible speculation.

If we dont do that and let thousands of millions of dolares be lost forever or not, that uncertainty will put bitcoin in a permanent speculation, but if you tomorrow delete genesis block of 10 years ago Satoshi Nakamoto millions and give them to the minning rewards, bitcoin would become more descentralized and out of that type of speculation.

If Satoshi is alive and listening he is in control, he have power to manipulate the market, lets put bitcoin resistant to that power, resistent to that trojan horse of speculation and show that bitcoin is a monster that even the creator cant put it down!

[Pmalek]

@cfbtcman
If Satoshi is alive, whether he is one person, a group, an organisation or a government, he had the chance to manipulate the Bitcoin market when it was at its ATH. Why do it now or sometimes in the future?

[aliashraf]

For some unknown reason I've missed this topic.

Generally speaking, there are two issues mixed-up in this discussion: 1- lost bitcoins problem, 2- blockchain bloat

Although the two are related but each deserve separate discussion. OP tries to figure out a way to solve problem one but to do this he finds himself faced with historical data problem and instead of focusing on this more crucial one grabs the very first ideas that have come into his mind while it is a more critical and important issue compared to lost bitcoins problem that he has started with and deserves more cautious approach.

As d5000 has cleanly discussed it, UTXO commitment technology is the most interesting approach to the second more general problem and I've discussed it in many occasions.

Reaching to a consensus about the state of the ledger in a given time (height)  is the most canonical definition ever for bitcoin. It is what bitcoin has been designed for and will always remain to be compatible with. It is why I think UTXO commitment is compatible with bitcoin and will be adopted in bitcoin and all its clones sooner or later. Without such an improvement bitcoin is exposed to centralization threats more and more because the pace with which communication/computing/storage infrastructures are developing are losing the war to blockchain growth

The problem with lost bitcoins is not a small one like what many authors has treated it in this topic and elsewhere. If we consider bitcoin as a system that is supposed to last like for centuries, it will eventually vanish because there is always a chance for some bitcoins to be lost due to many factors.

To fix this issue, there is just two options: 1- re-considering the 21 millions cap and 2- identifying and recycling lost coins.

The first option is out of context and OP is suggesting utxo age (not his term) as a measure to identify such coins. Once they are identified it would be easy to recycle them e.g. the protocol may adjust inflation rate to keep 21 million cap valid.

As of the measure to be taken, OP is suggesting like a century of age, I'm thinking of a decade, others may fall between the two. In any case 'removing' such utxos identified as 'lost' is where we need an infrastructure like UTXO commitment mentioned above. OP's proposal of "ignoring" blocks that hold no valid (young and unspent) utxo is not bad imo, but not as solid as UTXO commitment, it needs a lot of efforts and technical considerations with limited application unlike UTXO commitment which is useful for a much wider series of problems and at the same time an order of magnitude simpler to implement.

 

[cfbtcman]

@cfbtcman
If Satoshi is alive, whether he is one person, a group, an organisation or a government, he had the chance to manipulate the Bitcoin market when it was at its ATH. Why do it now or sometimes in the future?

Look Craig Wright, after he said bitcoin would go down it starts to go down, maybe because a fork war.

Maybe the guy dont like the path bitcoin is following or maybe because now he needs money.

When rich people starts to waste money they waste more and more, why he should mantain the money forever and die without moving it?

In that way of think why NASA would waste millions of dolares in a anti-meteorite plan?

If our civilization dont died until now its because there is no risk, look the dinossaures!

But this is not only a Satoshi dead or alive question, this is a technical solution that bitcoin needs sooner or later, so why not to do it as soon as posible?

[cfbtcman]

For some unknown reason I've missed this topic.

Generally speaking, there are two issues mixed-up in this discussion: 1- lost bitcoins problem, 2- blockchain bloat

Although the two are related but each deserve separate discussion. OP tries to figure out a way to solve problem one but to do this he finds himself faced with historical data problem and instead of focusing on this more crucial one grabs the very first ideas that have come into his mind while it is a more critical and important issue compared to lost bitcoins problem that he has started with and deserves more cautious approach.

As d5000 has cleanly discussed it, UTXO commitment technology is the most interesting approach to the second more general problem and I've discussed it in many occasions.

Reaching to a consensus about the state of the ledger in a given time (height)  is the most canonical definition ever for bitcoin. It is what bitcoin has been designed for and will always remain to be compatible with. It is why I think UTXO commitment is compatible with bitcoin and will be adopted in bitcoin and all its clones sooner or later. Without such an improvement bitcoin is exposed to centralization threats more and more because the pace with which communication/computing/storage infrastructures are developing are losing the war to blockchain growth

The problem with lost bitcoins is not a small one like what many authors has treated it in this topic and elsewhere. If we consider bitcoin as a system that is supposed to last like for centuries, it will eventually vanish because there is always a chance for some bitcoins to be lost due to many factors.

To fix this issue, there is just two options: 1- re-considering the 21 millions cap and 2- identifying and recycling lost coins.

The first option is out of context and OP is suggesting utxo age (not his term) as a measure to identify such coins. Once they are identified it would be easy to recycle them e.g. the protocol may adjust inflation rate to keep 21 million cap valid.

As of the measure to be taken, OP is suggesting like a century of age, I'm thinking of a decade, others may fall between the two. In any case 'removing' such utxos identified as 'lost' is where we need an infrastructure like UTXO commitment mentioned above. OP's proposal of "ignoring" blocks that hold no valid (young and unspent) utxo is not bad imo, but not as solid as UTXO commitment, it needs a lot of efforts and technical considerations with limited application unlike UTXO commitment which is useful for a much wider series of problems and at the same time an order of magnitude simpler to implement.

 

Good points!
About solutions 1 and 2 for fixing solutions i can imagine the 1st and 2nd together!
21 millions its to little, its not "human" money to look for decimal parts, and we shouldnt need FIAT money to calculate it, if someday with 51% user consensus network the number of bitcoins of each address was multiplied to the BTC unit comes closer to dolar and euro that would be a good marketing move and nothing would change in the real value.