Add new transaction type "unspent" to allow burned coins to be retreived.

[ghost]

CoZ asked for enhancements, soI'd like to present an enhancement worth roughly $150,000.

This transaction type would be identical in execution to a Contract transaction, signed by the sender, with one exception: The amount in the output would be subtracted from the destination address, rather than added to it. This would allow for people to reclaim coins sent to a "bad" address, but would not allow people to reverse transaction arbitrarily.

As you know, in order to claim gas you must "spend" NEO, send them to yourself. Any transactions that have been "spent" would be ineligible for "unspend". If a user wishes to prevent their coins from being "unspent" from under them, they either claim gas, or send them to themselves, confirming that their address is valid. this can be added as a default task in all wallets, which will also lead to smaller transaction sizes as you don't ever have more than one unspent tx per address.

Why is it worth $150,000?

There is currently $150,000 USD of NEO in burned accounts due to a ledger usb bug that displayed the wrong "spend" address.

AU3DGUrwAMF3rbc8mW8BrtWRsbBdBH5V4q: 1,947 NEO ALKfPGkaMVyVQjbv3Qa5bb6fWEURHmFr4S: 13 NEO ASHq4jJRgVTX2auvbDhAu8g4JQk2srV36D: 108 NEO AGAmSMhDkoJ2Wr7T4PeRHEf52eJJjzFfTK: 10 NEO

2078 NEO, $145,226 USD

https://github.com/CityOfZion/neon-wallet/issues/524

https://docs.google.com/spreadsheets/d/1afBgZ5yvr6FHN1CBl2bSlJIEfbpNtHukQSqW731gXmE/edit#gid=1139885301

[erikzhang]

So how can we tell which addresses are bad addresses?

[OrEagle97327]

Coranos' title refers to "burned coins" while I think of them as "locked" or lost.

My Neo is sitting on the block chain. ...I can document how they got there. Anyone can view them. No one can access them. Are they mine? Do they belong to the public domain? Do we consider this a "bad address" as questioned above? What are the rules for "reactivating" "locked" crypto?

The list of losses above, is very minor, when compared to this same concept across all crypto currencies. A tool like this could save a lot of broken hearts, and disappointed users.

What does the Crypto community feel should or can be done, with unclaimed and unusable crypto on the block chain(s)? What can be done for the good of all?

If we keep asking the questions, the right ones will get answered.

It takes about 10 years for a concept to catch on and become popular. (telephone, electricity, automobile, internet, email, facebook, and now crypto is 9 years old)

In the early 80s the personal computer became popular. In the early 90s the concept of "plug -n-play" came into use, sure made the computer user experience easier, and created widespread popularity. Until we develop a concept like this in the crypto world, it is our job, as pioneers, to leave a smooth path for others to follow.

[ghost]

There is no need to determine if an address is "bad" or not, the question is not relevant, as I will illustrate below. It will sound very strange, I just hope you understand and agree with me.

Every contract transaction that has not been spent, and is not an "unspend" transaction can be acted on by "unspend".

So, for example, this transaction could be "unspent" https://neotracker.io/tx/90551a4b7c941f105be2ecd186b79e640a06d475b834990edaeeb473756e1e0a sending 35,000,000 back from AQVh2pG732YvtNaxEGkQUei3YA4cvo7d2i to AWHX6wX5mEJ4Vwg7uBcqESeq3NggtNFhzD.

How could AQVh2pG732YvtNaxEGkQUei3YA4cvo7d2i secure his NEO from an "unspend"? Send all his neo to himself.

for example, this transaction cannot be "unspent": https://neotracker.io/tx/31819bb88636869e94cd6e8f5125c738d60ded4ecd91aeaa03a1b9ae5c563ab0

AXSoNQEKjmqPBNPg5cNrHyWivfjok3Vj9D sent 11,198 NEO to themselves.

Even if you did "unspend" that transaction, the NEO would go from AXSoNQEKjmqPBNPg5cNrHyWivfjok3Vj9D to itself, and then be unavailable for another "unspend".

This is similar to the way the NANOS coin handles their block lattice (which is where I got the idea), they require two transactions, a "Send" and a "Receive". This would require everyone's wallets to automatically "spend" any new inputs, to prevent an "unspend". That's the tradeoff, you can have reversible transactions, but need people to log into their wallets to receive funds.

You could add a 90 day timeout period if you wanted to, in order to give ample time for people to check their balance.

[erikzhang]

It can be implemented as a smart contract very easily.

[ghost]

I'm aware that NEP tokens can be reversed. This is about NEO, which cannot be reversed. In addition, ONT tokens were airdropped to these addresses.

Also, numero41, patience. This is the first time in the history of the world that we have a chance to do something like this. The chance to do something great. Don't let your passion overcome your reason. Emotion, yet peace.

[ghost]

Getting a reversible token, so the average person who is not technical can be safe from typos and random hardware faults.

As of right now, a single wrong keystroke can cost you .. everything. Or keep the private key in a file on your computer, which can also cost you everything if you get a virus.

Nobody in my extended family is going to use crypto as a method of payment if it's that brittle. They don't type that well. (This is true, when asked about crypto over the holidays, my advice was to stay out)

[ghost]

Thanks DL, I saw it earlier today.

Please keep the conversation to NEO, as this is the NEO github. There are many companies involved in this problem, but we can only fix it if we focus on each individually.

I don't want this incident closed because we kept getting off track, like the last one.

[roblee220]

I have the same problem, basically a missing MEW wallet whilst using the nano ledger s, I correctly went into MEW through the nano, sent tokens amounting to roughly 2000 dollars and now the nano is not showing the address and I obviously have no access to the private key, I’ve received 1 email from nano support in a week and I’d very much like to retrieve my coins as they are quite a large portion of my portfolio. I’ve sought advice for this issue but had no luck retrieving them yet what’s our course of action dies anyone have any ideas how to get these tokens bank?

[kryptob]

Good day,

I am Indeed one of the early victims of this issue and glad all attempts are being made to fix it. In August when Neo released the 0.03 wallet version and the announcement was made about how to earn gas, I transferred a total of about 319 NEOs to my wallet address. Before this I transferred 2 Neo to test it and earned gas for 2 days. The wallet froze at some point, I logged out and lo and behold I could not access the wallet again, till date it’s still inaccessible. That is roughly about $26K of today’s worth. As an early investor in NEO, this experience has been heartbroken till date. After several attempts contact COZ from August till c October that have not been any solution. I still see the Noe on neotracker and it’s still unspent, sitting there.

Apparently, when I tried to log back in, it had created another key. I have all the proof from the exchanges I transferred to and fro. Can this new feature help in any way please? I’m desperate to have my coins back.

[numero41]

@kryptob

Were you using the Ledger Nano to make those transactions?

About this new feature, Coranos knows much more about it than me as he implemented it, but I can surely say that it'll avoid this in the future, but won't reverse the previous "corrupted" transactions

[kryptob]

Hi,

Thanks for your response, ledger Nano had not been released then. I was using the 0.03 version of the desktop wallet. It was after the issues that COZ released a quick update to the wallet but I’ve been unable to retrieve my coins till date. Infact the only access to the wallet was using private key. It was after we complained (it happened to a few of us back then) that an improvement was made to the wallet to include other routes e.g. passphrase , passwords of accessing it. It happened in August 2017, my coins are still there untouched so this is no scam or anything of such. I have all the proof of ownership as I moved them from exchanges and can provide any necessary information.

[numero41]

We all have all proofs and necessary informations you mention, but it won't help you in any way.

For now we never had any information that could let us hope for a way to gain access again to the wallets. The only hope would be to find a kind of responsibility of some of the parts involved in the transactions.

This crypto world has no regulations yet, and for now not anyone (wallet developers, Ledger, exchanges, etc) can be taken as responsible for any loss if you can't prove that you didn't make a user error.

The only way for any of us to have any kind of help/hearing, is to be able to reproduce the "bug", so that the source of it can be identified.

Once it'll be identified, actions can be taken to find who's responsibility it is, and how to get leverage to have a compensation.

But once again, I would avoid kind of hope if I were you (I lost about the same amount as you did), which doesn't prevent you from continue to search, raise awareness, and seek for tech people that could give a try to find what happened in exchange for a compensation (like 20% of the funds)

[kryptob]

For me it is only logical that for lost coins, they can burn those coins using smart contract and subsequently mint new ones as replacement to the address they were sent from or can be reclaimed by the legitimate owner if proofs are available. What are the specifics of the features Coranos is working on?

[numero41]

If your private key doesn't derive the correct public key anymore, you can't prove you're the owner, that was my point.

What kind of proof are you talking about? Transactions records from your exchange account history + blockchain explorer txids? That's not enough!

Let me give you an example : You have a very good friend to who you want to send some coins, he gives you his wallet address, you make the transactions, and you have all of them recorded on the exchanges, and blockchain explorers.

Then you have an argument with this friend, and want your money back, but he wouldn't give it back.

In that case, you could still come here (or any kind of support), and claim you made some transactions from some exchanges to your wallet, but that unfortunately you lost access to it (like it happened to all of us). There would be no way for anyone to know if you're saying the truth (that you were the initial owner of the wallet), unless you have the correct private key that derives the correct public one. And in that case, you wouldn't be here at all lol.

That's all the perversity of what happened to us...

[ghost]

Nobody is currently working on this feature request. They don't seem to care, regardless of how much proof there is.

I do not know how to get them to care, as I have tried all I could.

[kryptob]

@numero The case you mentioned is an outlier. There’s some level of proof if one can confirm that your stuck coins emanated from an exchange account you own, that is often times near accurate that you owned the coins. The case you eluded to is exceptional.

@Coranos, is this feature possible technically at all? Where they burn the stuck coins and mint another in replacement for you as long as you can proof you own them. Since it’s open source, can any developer build this feature and have COZ commission it. If I’m not mistaken, the total value of lost coins/ stuck coins should be nearing $1m mark if not more.

[ghost]

The feature is technically possible. As said above, it's actually very easy to do as a smart contract.

However you would need to convince NEO to run the modified code on their consensus nodes, or there's no point. I do not think it is possible to convince NEO to run the modified code on their consensus nodes.

Their original whitepaper mentioned the feature (reversible transactions), but it is currently absent from the whitepaper and the code. I didn't save a copy of the whitepaper, and the link is dead, so I have no proof.

[numero41]

"There’s some level of proof if one can confirm that your stuck coins emanated from an exchange account you own, that is often times near accurate that you owned the coins."

I want as much as you to get my coins back!

Can you give me ONE example where someone "proved" its ownership to a wallet using this argument? I wish there were... In that case we all have our proofs ready to go.

[kryptob]

''Their original whitepaper mentioned the feature (reversible transactions), but it is currently absent from the whitepaper and the code. I didn't save a copy of the whitepaper, and the link is dead, so I have no proof.''

I kid you not. There were about a few of us affected, we saw the feature in the white paper as well. I think one of the guys saved it, it was on the Antshares, then on Neo when it was rebranded. After our complaints, they removed that section and replaced the whitepaper. They specifically stated that transactions as such could be reversed.

[kryptob]

@numero41 i agree with you, proofing ownership from a wallet is hard, but if you can prove it on an exchange that is easy. The transaction IDs are recorded, the wallet address where it emanated from are also there and i'm sure support in exchanges will be able to help as well.

[numero41]

@kryptob

You know, that was the first thing I told myself, 4 months ago when it happened to me, that I had all the necessary informations and that no one could contest it, I have my exchanges accounts, which I am the only one to access for sure, the txids, everything that shows that I for sure did made the transactions this exact day, at this exact time, with those exact amounts, etc.

Then I was told almost immediately from everywhere (in answer to my first posts) that I wouldn't get anything looking in that direction, so I didn't event contacted the exchanges about that.

Anyway, as the transactions can't be reversed, the only way to get our money back is to find some part that has to be taken as responsible for the loss, and willing to give the same amount in compensation for a malfunction, right? What exchange would help in any way in that direction?

On your side, did you make any try about it?

[OrEagle97327]

@coranos Who in the Neo team would be most likely to help you implement your idea in this thread? Who else in the community has shown any interest?

What additional venues would be good to pursue a campaign to push for this?

Imagine how the whole crypto community would be with this as a standard feature.

[ghost]

@erikzhang would be the only one I know who would be able to implement Unspend transactions.

Most times I mention reversible transactions, people are hostile to the idea. "Caveat Emptor" as they say.

[OrEagle97327]

We are looking for something between "Caveat Emptor" and "Plug and Play"

I do not understand the reason for hostility.

fix my oops; ...Seems like a reasonable feature.

[shargon]

Im sorry, but if it has been lost, is lost .... otherwise, it would break the main principle of a blockchain

[OrEagle97327]

Good point... ...re Block Chain

Not lost... ...Just unavailable (Can be seen by block chain viewer)

Just like natural resources on a new frontier. How do we mine that?

Imagine how much value is setting on all the various block chains. Can the community come up with a form of reclaiming.

[grishatataru]

So whats the next step? How can we get this issue fixed? Im having the same issue,my public NEO address has changed,when i log in using Ledger Nano i see 0 gas and 0 neo. Showing UNSPENT in neo tracker ANy help???? AVjDZ7VVz7foeqFmdW2R5okbjr1pTEmEBs

419 NEO

Sent to

AcWyPgfSVovTnE5rG6w3KNMCcoipjuDBxP 419 NEO

[grishatataru]

How can we get this issue fixed? Im having the same issue,my public NEO address has changed,when i log in using Ledger Nano i see 0 gas and 0 neo. Showing UNSPENT in neo tracker ANy help???? AVjDZ7VVz7foeqFmdW2R5okbjr1pTEmEBs

419 NEO

Sent to

AcWyPgfSVovTnE5rG6w3KNMCcoipjuDBxP 419 NEO

[grishatataru]

Im willing to give 100 Neo to someone who can help

[ghost]

@erikzhang since we have seen a resurgence of burned NEO, can reversible transactions for NEO be implemented?

[erikzhang]

I'll try to design a mechanism to retrieve lost assets. But this is a long and low priority job, so please don't expect it.

[grishatataru]

Thanks a lot

Sent from Yahoo Mail for iPhone

On Thursday, July 26, 2018, 9:27 PM, Erik Zhang notifications@github.com wrote:

I'll try to design a mechanism to retrieve lost assets. But this is a long and low priority job, so please don't expect it.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[vncoelho]

I believe that most part of the fault on this reported problems was not Neo core library, but faulty on: maybe on the Ledger Hardware; maybe on the Ledger software; or maybe Neon wallet incompatibilities (all of them are third party software's).

I am really happy to read @erikzhang last message, that there is this consideration of implementing a decentralized mechanism for recovering lost assets. You all could also count with me on designing that.

Another possibility that come into my mind now maybe could be something to recover during NEO 3.0 upgrade to native contracts. But, as Erik said: "this is a long and low priority job, so please don't expect it.".

[ghost]

It could definitely be a problem with the ledger, or any of the other moving parts.

More coins than just NEO had it happen. All the software involved is under a limited liability license so it doesn't really matter who cause the problem, nobody has liability.

I still think it would be a useful feature to be able to unspend coins in cases where it is obvious the spend was faulty. Blockchain purists say this breaks a fundamental aspect of the chain. But widespread adoption won't happen if yo can accidentally lose everything, with no method of retrieval.

[vncoelho]

@coranos, maybe the focus now should be more related to native contracts since global assets as UTXO are planned to be migrated to native contracts (PR #639).

[igormcoelho]

@erikzhang perhaps I have an idea. Since Hash160 is an unreversible and unpredictable function, if someone manages to understand the algorithm that lead the the address failure, we can reverse these tx without breaking any principle (nor dangering no one's assets). For example, suppose the wrong function has added 1 to the scripthash (I don't know if that was the case!!!), no one would ever be able to generate a private/pubkey to match the expected script. So if, by the way, someone manages to do that, we can create a NEP that allows address on this "bad" format to be recovered. This is just a feeling, first thing would be to understand how the failure happened... and if it was a random one, it would be nearly impossible to actually know if these were correctly spent or wrongly spent. Blockchain is all about following rules, and the value of the assets are based on their security, utility and scarsity. So if any change can be made to the network, without breaking any of these things, it could be a good thing afterall.

[kryptob]

Hi,

What if one moved Neo from an exchange (who will have the transaction history, to a faulty wallet and the funds are still unspent till date. Is there a possibility of reversing that transaction?

On 9 Apr 2019, at 20:35, Igor Machado Coelho notifications@github.com wrote:

@erikzhang perhaps I have an idea. Since Hash160 is an unreversible and unpredictable function, if someone manages to understand the algorithm that lead the the address failure, we can reverse these tx without breaking any principle (nor dangering no one's assets). For example, suppose the wrong function has added 1 to the scripthash (I don't know if that was the case!!!), no one would ever be able to generate a private/pubkey to match the expected script. So if, by the way, someone manages to do that, we can create a NEP that allows address on this "bad" format to be recovered. This is just a feeling, first thing would be to understand how the failure happened... and if it was a random one, it would be nearly impossible to actually know if these were correctly spent or wrongly spent.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

[ghost]

This all depends on how you define a 'faulty transaction'.

From my research last year the only way you could tell a faulty transaction from a real one is the faulty transaction never claimed gas.

So your choices are: 1) Have a central authority who can reverse transactions, where you submit a request and they perform an unspend transaction. 2) Allow anyone to reverse a NEO transaction up until gas is claimed, which locks the transaction.

Any third suggestion seems to be with 1) or 2) with extra steps.

[vncoelho]

I like @igormcoelho suggestion. In particular, it might be possible for the cases of the Ledger reported in this issue. Maybe someone could investigate the reason of that faulties and try to report the possible algorithm that was executed behind and generated the faulty addresses. If that algorithm could be reproduced it would be a great step toward a proof and a mechanism for recovering such funds. @coranos, since you are one of the contributors of the NEO app on Ledger maybe you could try some insight about that cases reported in this thread. Following Igor's suggestion might be a good path.

@kryptob, the other cases are quite special and I think we need much more time to think about them. But we never know the future. Let's try to create something together.

@coranos, centralized agent does not currently exist in NEO (it is a multi-agent based protocol). In this sense, option 1. may be discarded.

[ghost]

There is no way for me to go back in time a year and a half to that hardware and software version and reproduce a problem that happened on 36 of ten million ledgers. I tried to reproduce the problem for over 80 hours from December of 2017 to April of 2018, and did not encounter the problem a single time. This is not a productive area of research.

The error rate is small enough it could literally be cosmic rays or USB bus failures. Hardware problems that would not exist on any but the faulty host's hardware.

I have not had a bug report since August 2018, so I assume ledger 'stealth fixed' this issue and will not be cooperative in researching an issue with potentially millions of dollars in liability and bad press. I asked. The email received no reply.

Centralizing agent does exist. Every time the blockchain stopped they simply distributed a new database that 'got past the stuck block'.

But if you want to ignore that and discard 1) then 2) Is your only option.

[vncoelho]

Ok, @coranos, I understand the difficulty and appreciate the efforts you already did over this, as well as the NEO app on Ledger.

About the centralized agents, I think that you confused the concept. The distributed database is a way to facilitate the synchronization. You do not need to use that database to sync your node. There are plenty of ways to sync a node without that database. Furthermore, NEO has block finality and two branches never existed. Just duplicated blocks that stalled clients, an issue that was recently solved with a notorious scientific advance over the topic now with dBFT 2.0.

I need to study, research and investigate more to think if 2) is really the last option.

[ghost]

So, uh, you will still get a stalled blockchain periodically.

MinerTransaction only has a nonce passed into it's hashing function. So you will eventually make another mining transaction which has an identical hash. Chances increase as you mine more blocks. Current chance is (3,600,875 in 4,294,967,295) so about once per thousand blocks.

When that happens, you have to centrally restart the blockchain. Happened about six times last year before I lost track.

The code line i'm talking about.

https://github.com/neo-project/neo/blob/38cac283c737de62d3631ad39f97b3c59ea8969f/neo/Network/P2P/Payloads/MinerTransaction.cs#L11

If it's fixed somewhere else, nevermind, I havn't looked in the code in a while.

Like Eric said, It's a hard problem.

[vncoelho]

I did not understand the correlation of this last point with what we were talking, @coranos, sorry. Maybe it is better to open a new issue for this discussion if you want to clarify that and discuss this possible impacts of the Minner transaction and any clarification regarding the Consensus MAS algorithm Otherwise, we gonna run out of the scope of this thread. There is also the Technical Specification (Yellow Paper). The dBFT chapter is almost complete, it may provide new insights about how dBFT has been running: https://docs.neo.org/en-us/08_dbft.pdf

[igormcoelho]

So, uh, you will still get a stalled blockchain periodically.

Int nonce is not good entropy, but from my understanding MinerTransaction won't exist anymore on Neo 3.0, as everything will be Invocation. So, certainly it will be necessary to add more entropy to this random, and at least to adopt to cryptographically safe random generators (of much larger scale than 4 or 8 bytes, let's take 32 for example), so I'll do the issue now to not forget that: https://github.com/neo-project/neo/issues/697

As it's impossible to detect what issue caused the problem (it could be cosmic rays indeed), we should improve this in many ways: guaranteeing that the software/hardware won't do a mistake (we need to verify this transaction output somehow), and test it too (I miss testing capabilities on Neon wallet, when I just create addresses I'd love to automatically test it to make sure it works). For proposal 2, for me all rules are acceptable on the blockchain, as long as they are public and well-known. For example, if a rule exist that funds "stalled" for over many months can be reclaimed, I won't argue against it. As long as this time is not too short, it won't affect business. I fully agree with that, blockchain is not near the current needs of regular users (even myself, I'm always scared!). This feature would be really, really important. You can count on me @coranos .

[erikzhang]

I have thought about a solution to this problem before.

If the private key of an address is lost, it means that no one can transfer the token from this address. So if the token on an address has not been transferred for a period of time (for example, one year), then we can think of the address as the "possible loss of private key" address.

For a "possible loss of private key" address, the user can send a transaction to request the withdrawal of all assets in the address, but must pledge the same amount of assets in the transaction.

If, for a period of time (for example, 5 years), no one has objected to the application with the correct signature, the assets can be transferred by the new owner. And previously pledged assets can be returned.

But if someone provides a signature to prove that he is the owner of the address, then the user who sent the withdrawal transaction is a malicious user. Then the assets pledged by the malicious user will be confiscated (Or directly forced to reward the user who provide the correct signature).

[shargon]

If the private key of an address is lost, it means that no one can transfer the token from this address. So if the token on an address has not been transferred for a period of time (for example, one year), then we can think of the address as the "possible loss of private key" address.

Or maybe is a Holder :P

For a "possible loss of private key" address, the user can send a transaction to request the withdrawal of all assets in the address, but must pledge the same amount of assets in the transaction.

What user? any?

The problem with this, is that we will have a lot of bots doing this job, waiting for request, if two users made a request, who win?

If is lost, is lost... 🔒

[erikzhang]

The problem with this, is that we will have a lot of bots doing this job, waiting for request

I have provided a way to prevent malicious applications in the solution.

[erikzhang]

If is lost, is lost...

I also agreed to this. I proposed this solution just for discussion.

[ghost]

how about withdrawal can only go back to the address it was sent from. 'unspend' implies they 'send one step backwards'. if it's a holder, they would claim gas I would think.

[kryptob]

Is this a possibility ?

Resend back to the address it was sent from ?

how about withdrawal can only go back to the address it was sent from. 'unspend' implies they 'send one step backwards'. if it's a holder, they would claim gas I would think.

On 10 Apr 2019, at 10:44, coranos notifications@github.com wrote:

how about withdrawal can only go back to the address it was sent from. 'unspend' implies they 'send one step backwards'. if it's a holder, they would claim gas I would think.