Bitcoin Forum

Solana does it. Didn't know they did it when I first started using it:

https://docs.solana.com/implemented-proposals/rent

Short answer: we don't need it. But even if you want to do that on Bitcoin, you still can.

What is the purpose here? To have some data pushed to the blockchain? If so, then you can use OP_RETURN and pay for every pushed byte. You can also place your OP_RETURN in some unused TapScript branch, tweak your Taproot address, and then pay no fees, but keep the proof that your commitment is connected with your address.

It seems that case (2) can be handled by case (1) if you use Lightning Network, then things are much simpler. You can just create some channel and create channel closing transactions upfront, so that one side will get more and more coins, just by waiting and publishing a transaction with some higher locktime.

No problem with that, the latest closing transaction can move all coins to the one party, then that person would "collect rents from stale accounts" (simply because another party did nothing for the whole time since channel creation to the locktime in the last channel closing transaction).

In Bitcoin, you don't need "the whole scan of accounts", you only need to check accounts that voluntarily joined your system, for example by forming channels with you (or joining your channel factory, N-of-N multisig with Taproot is cheaper than it was with other address types).

Is it user-destroyed or miner-destroyed? Because using any user-destroyed method is going to leave a trace in the blockchain. On the other hand, miner-destroyed way is much cleaner, because then such miner is just taking less coinbase reward, so there is no additional UTXO at all (and this can be enforced in a no-fork way, where only a group of miners will burn their coins, while still being part of the network).

Wait, there are "validators"? So why this altcoin is not directly connected with Bitcoin by a two-way peg-in? And why this altcoin needs any new coins at all and not reuse Bitcoins? If you have some fixed group of "validators" that are allowed to mine, then you don't need any new coins, you can run everything directly on Bitcoin blockchain, by creating a signet sidechain, pegged into the mainnet! Even more: if that group of "validators" can evolve over time (so that introducing new "validators" or changing existing ones is possible), then it can be easily handled, just by changing keys and signing that! So, that kind of staking model can be running on Bitcoin today, without any forks!

At that point, I stopped reading, because now I am convinced that this system can run directly on Bitcoin, so we don't need to change anything to allow rent.

I would say it's not good idea/not necessary for following reason,
1. Most people deemed current node amount is still sufficient. At time of writing this post, Bitnodes report there are 15113 reachable nodes[1] while Luke Jr stats (which include node which doesn't enable incoming connection) report there are about 50K nodes[2].
2. Increasing complexity of Bitcoin protocol.
3. Add more burden for Bitcoin contributor (whether it's on protocol or implementation level).

[1] https://bitnodes.io/dashboard/ (https://bitnodes.io/dashboard/)
[2] https://luke.dashjr.org/programs/bitcoin/files/charts/software.html (https://luke.dashjr.org/programs/bitcoin/files/charts/software.html)

I'll skip my comments and focus on this:

Wait a sec, isn't this an absurd idea? Why should I require money to get self-sovereignty and privacy? Aren't those enough to cover my disk's and memory's expenditures? It sounds like they want to expand their network in the wrong way.

I dont necessarily like the idea of paying an involuntary rent, but if your balance is over a certain amount it's "rent exempt" and that amount is not too large it didn't sound like. So basically what rent is doing is taking very small accounts and getting rid of them over time if they go unused and keep a small balance. What is troublesome is the fact that once an account would get to zero on solana i think they delete all the accounts data including any nfts that were owned. how to get those back? not sure if it is possible.

anyhow, for bitcoin, paying rent for small balances could provide an extra income to miners i suppose to help offset the decreasing block rewards. not saying i'm for it though.


#2 does seem like a reasonable objection. i just realized that implementing rent on a utxo based system like bitcoin is not so straightforward as it is in an account based model like solana because you don't know what addresses make up an account. so you would have to treat each address as its own account...but if you did that, it helps get rid of small utxos that take up the same space in the utxo set as larger utxos i guess it would encourage people to consolidate their utxos at the very least  ;D

There is no reason why nodes can't collect money in exchange for providing data to others. Furthermore, it can be done without any changes to consensus rules.

It goes against the nature of the peer to peer network where everyone is a peer at equal level which means any money charged would be negated (peer A charges $1 to answer peer B and peer B charges $1 to respond).

But, everyone is not an equal peer. Lite wallets, SPV wallets, and surveillance companies use the data that nodes provide, but contribute nothing.

And as you note, the charges would balance out anyway for those who are equal peers. Last month, my node downloaded 1GB but uploaded 10 GB.

But we can't consider SPV clients part of the network in first place, the network consists of full nodes and everything else is a leach not a peer.

Exactly, and they would be the ones that pay for the data.

Anyway, I agree with you to some extent. The incredible bounty that has been provided to the world by open source software developers is greatly under-appreciated. But, where would we be if they had charged for their work?

But, the reality is that nobody can dictate how any issue will be resolved in a decentralized environment. So, fee or no fee -- we will just have to wait and see.

Where are they?

Right?


then that means you're giving more than you're getting. seems like people (nodes) that are giving more than they are giving should receive compensation on their net contribution. [REALLY gets on soapbox] miners arent the only ones that contribute to the network being functional. but if something like that was implemented, it would have to be robust against people trying to game the system. but it seems like bitcoin only rewards miners and not nodes that don't mine but still have costs for helping the network. kind of unfair in my opinion but anyhow...


Imagine that! Users that contribute nothing being required to pay for something. Sounds about right. ;D



Maybe nodes could have the ability to tack on their own transaction fee... :D

Since Lightning Network, it also rewards online nodes, willing to put their coins in channels to provide liquidity.

I don't think so. They will just start running their own nodes (or fake nodes, just to get some reward). It will be the same case as with the Lightning Network: there are wallets charging more satoshis than people would pay on-chain. People will not pay high fees, people will optimize their costs by running their own nodes or switching to another, cheaper network for a while.

Technically, they can. Practically, they have to do it in the right way, because in other case, people will react and optimize their costs. Also, if you have some Lightning Network wallet, like Phoenix, then they charge 3,000 satoshis for on-chain operations, just because they can. Also they charge 1% of coins. That simply means, after reaching 300,000 satoshis, it is more profitable to withdraw coins on-chain than to form a bigger channel. See? The problem is not technical, the problem is about total cost for the user, depending on what that user wants to do and how many payments there are.

i would think the percentage of people running a full node that also run a lightning channel is not so large. maybe 5 or 10 percent if that. as well, LN is layer 2. i'm talking about layer 1. could the network function without full nodes that arent doing mining? if not then it's hard to see why they arent rewarded in some way. presumably they do it because the reward outweighs the associated costs but the reward is not on chain. it's something less tangible.

It would be much slower in terms of bandwidth. I suspect the mining nodes are very few if we compare them with the tens of thousands of full nodes[1]. Note that miners don't always run their own full nodes; this is a requirement of the pool. Those who participate just hash block headers.

Again, they are. They gain the ability to verify everything. They need none, which also enhances their privacy.

---

[1] https://bitnodes.io/dashboard/

They are rewarded in many ways but they aren't getting paid if that's what you are looking for.
The reward is a higher security for both the person running the full node and by contributing to the network, more security to the only decentralized currency in existence...
It is is also the improved privacy that running SPV clients can't provide and obviously it doesn't even exist in any other client type.

a month is 4032 blocks (call it 5gb)
so you only downloaded 1gb

you should be downloading 1gb a week even from just 1 peer.
(1mb block*144 a day*7day=1008mb/1.008gb)

so your not even download all your blockdata.
heck i use 160gb a month just watching movies. and you worry about your litewallet/pruned/headers only data

why are people trying to make it sound like the bitcoin network is soo expensive when no one ever complains about youtube or netflix (1gb SD, 3gb HD, 7GB 4k.. for just 2 hours)

the reason why ethereum is moving forward and becoming more popular platform for multiple projects is because even though their blockchain is 1TB (3x bitcoin) they are not playing political games of pretending that blockchains are expensive/need compensating/dont work/shouldnt be used for daily use.. they are actually trying to make their network have more use

Yeah I guess more security but in reality, you can sign an offline transaction and broadcast it without needing to be a full node. Isn't that just as secure? Now about your statement that bitcoin is "the only decentralized currency in existence" that's a pretty big statement. I would be willing to grant that bitcoin is the biggest decentralized currency in existence but I'm not sure how one comes to the conclusion that it's the only one. :o


Once again, create an offline transaction, sign it, use a vpn or torr and broadcast it. How's that for privacy and all without the need to download the entire blockchain.

Oh, I know people might not look upon this type of behavior as being very "contributing to the network" and think of it as more of being a "leech" i get that because it is. So we do need people that stand for something out there running full nodes for sure.

And you sign that transaction based on... block explorers? Some data from SPV nodes? You fill your transaction inputs based on trust that those previous transaction outputs really exist. But in case of some attacks you could be convinced to sign coins that never existed at all.

Security is not limited to whether or not you leak your private key.
For example without a full node you can not be sure if your unconfirmed transaction is really unconfirmed or the third parties you rely on are lying to you. Or you can't know if a confirmed received transaction is your tip is still confirmed or double spent while you are stuck on a stale chain. That's just two examples off the top of my head.

I'd love it if you could show me another one!
The others are either centralized or semi centralized or simply they are not currencies.

Your wallet normally has more than one address in it whether it is HD or not. When you connect to a SPV server you are telling them all your addresses and whether your IP is real or not they still can link them all together.

While i agree with block size increase, your comparison isn't fair. Simply using few hundred GB/month is easier than verify and store it on your device.


Or attack which give you wrong balance of input which could lead to overpay TX fee.


How about Monero?

Monero is probably the closest project to being a decentralized currency and I don't really have any criticism of it but I do have concerns.
One of it is the scaling issues it has which is a lot worse that what bitcoin faces and it will make it hard for XMR to be a good currency.
Another concern of mine is how easy it is for them to push forks! The lack of resistance is not always a good sign, it could also be because of low node count and lack of distribution.

Another way that it it distinguishes it is the work that is done. There are uncountable of repositories related with Bitcoin, from Bitcoin Core and other clients to SPVs servers, payment processors, second layers, wallets, block explorers, libraries etc.

Then there are the books, the talks, the tutorials, the articles, this forum! Generally, there's a high degree of dedication which makes it easier for every newcomer to learn and build interesting things.

I recently found out that the community decided to change the mining economics (https://bitcointalk.org/index.php?topic=753252.msg59446380#msg59446380). I don't know if that is subject to decentralization, but it's a serious downside.

yes, all the big ones. if they all agree then that's almost as solid as running my own node I would think.


that's right. and the probability that they all agree and all are wrong would be neglibible as I check additional block explorers for confirmation.

and what would be the harm in doing that? once i realized the issue i could just respend on the correct chain right?

Well those things might be true but having access to a wide variety of blockchain explorers solves the issues of security as you define them I would think.


As someone else mentioned, monero. You dont think it is decentralized enough but it actually has a better situation than bitcoin with regards to its mining algorithm. It's closer to the idea of "one cpu one vote" than bitcoin that's for sure.  ;D bitcoin will never be that again. but monero will be. not saying monero is easy to mine but its mining algo is probably better at encouraging decentralization over the longterm whereas bitcoin's is just the opposite!
 
With that said though, things like Litecoin seem to be somewhat decentralized:

Litecoin was launched with the aim of being the "silver" to Bitcoin's "gold." Like Bitcoin, Litecoin is a peer-to-peer internet currency. It is a fully decentralized, open-source, global payment network. Lee developed Litecoin with the aim to improve on Bitcoin's shortcomings.

Some wallets I used were that way but it is possible to use a single address as your wallet. Sending change back to it, etc.


Don't worry about that. They push solid stuff out the door.

That's a naive conclusion. One might say that due to constant changing of the mining algorithm and lack of interest in the project, there hasn't been enough miners to grow the hashrate. There is a lot of smaller altcoins with low hashrate that you can mine with CPU!
This also increases the risk of 51% attack due to a much lower attack cost, even though such attacks are slightly different in a privacy oriented cryptocurrency such as Monero.

Litecoin is a weak copy coin that only changed block interval, total supply and mining algorithm to what they promised to be "ASIC resistant" which it clearly wasn't. All these changes are meaningless and without real innovation, the rest they just copy paste from bitcoin.

The concern is not about the quality of the stuff.

I think you are wrong and it is technically possible to decentralize mining. Some ideas are still in progress, but I think someone will come up with a solution, sooner or later. The easiest and the most inefficient is of course collecting all shares from all miners and treating 80-byte headers in the same way as transactions, but just on a different layer. But that doesn't scale, in the same way as blockchains doesn't scale, and you eventually have to process up to max_number_of_miners*max_block_size per each superblock, so it should be optimized or solved in a different way.

I don't think so, because the mining algorithm is not the problem. You could have decentralized mining on double SHA-256, you just need some way to solo mine a million times easier block and get a million times smaller reward for doing that in pure P2P way. Also, if you complicate the mining algorithm, then it is harder to verify blocks, you can clearly see that on CPU-based coins, where initial chain download is quite fast, but initial chain verification takes a lot of hours, even if the size of the chain is around 1 GB. You can run two offline nodes, one will have the whole chain and one will be empty. You can even use SSD and have everything downloaded, it will still take a lot of hours, just because hashing block headers will take a lot of time (and it will be a disaster if that coin also uses the same algorithm everywhere, for example to build a merkle tree, then checking 1000 blocks can be as hard as mining one block!).

There is no reason to harm your privacy, where you can use a brand new address each time (that could come from some HD wallet), and pay the same fees.

they used to have to change the algo constantly because of asics.

And yet litecoin is fast and has very low transaction fees. making it more attractive (than btc) in some ways.


ok then?


https://en.wikipedia.org/wiki/Bitcoin_scalability_problem
7 transactions per second is pretty bad it can't get much worse right?

at least monero has dynamic block sizes.

Or maybe it's just that you don't understand the monero ecosystem. They don't call them "forks" they call them periodic "network upgrades". for example in nov 2019 they upgraded the network with RandomX and a few other minor things. we might imagine that the userbase generally saw this as a very big net positive as they probably do with all such upgrades and  thus went along with it.

oh and by the way, the amount of "forks" as you call them has not been every 6 months like it used to be. it's alot less in the last few years.

I think we don't have to increase block size to increase transactions per second, because there are non-interactive ways to batch multiple transactions into one, making it smaller and serving more users with less bytes, and in my opinion that's the way Bitcoin should take.

It doesn't matter how they call them. If you are forced to upgrade every time, then it is bad for the stability of the protocol. Bitcoin is going to do hard-forks only when forced to (for example in 2038 the current version of Bitcoin Core will be incompatible, because of reaching 0x7fffffff time value and handling time in a wrong way during convertions, for now it is the only case that could need a hard-fork I can see).

i think you do.


how?


that all depends. i wouldn't say i agree with your statement though. because it is a sweeping generality.


bitcoin will do hard forks whenever someone wants to fork it. not when "bitcoin" decides to.


batching multiple transactions seems like putting people into a 2nd class citizen heirarchy with regards to their money transfer no thanks. bitcoin is already slow enough as it is. also you can't really predict the future and when and under what circumstances bitcoin might be required to do a hard fork.

For example by using things like MimbleWimble (https://bitcointalk.org/index.php?topic=5387856.0) or CoinPool (https://bitcointalk.org/index.php?topic=5389702.0), but I think more will come. People are creative.

Yes, you can always do some hard-fork, but if you are in economic minority, then you will end up with some other coin than "Bitcoin". There are many forks like BCH or BSV, but I think some useful features like SIGHASH_GROUP could be imported (like Schnorr signatures were), and the main difference will be just a block size, where Bitcoin will batch things in non-interactive way, when other chains will grow ad infinitum and call "scaling" something that does not scale. If you need 1000x bigger blocks to reach 1000x more transactions per second, that's not scaling, that's linear growth, where you need 1000x more resources, so you have the same scale.

In case of timestamps, it is possible to set your system clock to 2038 or 2106 and see that the chain will stop or the client will crash (https://bitcointalk.org/index.php?topic=5365359.msg58166985#msg58166985). So in this case I am sure that the newest version of Bitcoin Core cannot handle it here and now. So I know for sure that we would need a hard-fork in this case, unless we are going to do weird tricks, like replacing the latest block and pretending that the chain is still going forward. Maybe some hard-fork will be needed in other cases, but for timestamps I am quite sure about that.