(Ordinals) BRC-20 needs to be removed

[ABCbits]

You still can use Linux on old or low end machine, assuming you choose light DE (such as XFCE) or distro specially created for such machine.

yes but older hardware and older cpus are becoming deprecated by the linux kernel developers.

Linux Kernel May Drop i486 Support as Torvalds Backs Pentium Plan
https://www.tomshardware.com/news/linux-removes-486-cpu-support

Linux Kernel Nixes IDE Support In the Latest 5.14 Release Candidate
https://www.tomshardware.com/news/linux-kernel-nixes-ide-support-in-the-latest-514-release-candidate

It's far more than older hardware. i486 disconnected on 2007[1], while SATA (as successor of IDE) exist since 2003[2]. Although if you intend to keep using device with i486, consider NetBSD which claim to support i486[3]. Although FYI i386 which also disconnected on 2007[4] still supported by some Linux distro, such as Debian[5].

With just 8 GB RAM, Bitcoin Core is pounding my SSD at maximum capacity for hours during IBD. Even though it's just a one-time thing, it's only getting worse.

The solution to your problem is to get a more poweful CPU and alot more RAM. To be quite honest, that's the only solution to your problem. And maybe a faster SSD. Is your SSD NVME or is it SATA? Maybe you need to get an NVME one.

If you happen to own computer with big RAM which can store all UTXO, using HDD should be fine.

[1] https://en.wikipedia.org/wiki/I486
[2] https://en.wikipedia.org/wiki/SATA
[3] https://wiki.netbsd.org/ports/i386/
[4] https://en.wikipedia.org/wiki/I386
[5] https://wiki.debian.org/SupportedArchitectures

[BlackHatCoiner]

You seem to have suggested previously that ordinals is "good" because it stops people from using something like bitcoin stamps instead.

I have said that I, and all of us, would rather leave Ordinals as they are, than moving them to the UTXO set. I don't think anyone would disagree with this statement. Ordinal users would continue clogging up the network with trash, it's just that this time, we'd have an extra burden to account for; an inflationary UTXO set. Why should we incentivize ourselves towards that direction?

i disagree that something can be judged as good or preferred just because it stops or offers an alternative to something which is a built in feature of bitcoin (multisig) being used in way that you think might be harmful.

If something can't be judged as good or preferred according to you, then why are you arguing that Ordinals should be moved to the UTXO set?

[mikeywith]

The solution to your problem is to get a more poweful CPU and alot more RAM.

This isn't exactly cheap or efficient. Anything that relies on infinite resources is deemed to fail. In fact, if adding more resources were actually a good "fix," then we would never be talking about the UTXO set. The whole idea of the UTXOs set was never a part of the original Bitcoin software. It only came into existence when "adding more RAM" was no longer feasible. The original method of verifying transactions involved searching the entire blockchain, and then we hit a brick wall and had to find a meaningful solution, which was the UTXO set (not more RAM).

The ideal scenario for outputs is to never be stored on your disk. They should be in the memory buffer, then wiped out as soon as they are spent. But because the set has grown too much, we need to flush them to the disk. Obviously, having more of them needing to be in the memory buffer or even on disk makes transaction verification, block reorgs, and everything else more time and resource-consuming.

I'm still baffled how, and i know this might sound a bit off topic, but how downloading the entire blockchain is apparently some type of CPU/RAM/SSD-HDD intensive process. That makes no sense to me.

Glad you mentioned this; it's another issue with the UTXO set. The set is never downloaded; it gets extracted from the blocks. You need to download the block, verify it, extract the UTXO set from it, and move to the next block. The number of iterations you have to do is going to be similar regardless of which way you decide to sync your node. The more UTXOs, the slower the process. Even if you manage to magically have the entire blockchain copied to your drive in 3 seconds, digging into every block to construct the UTXO set is most certainly going to take forever.

Another point worth mentioning is that spending 2x on your RAM or anything else isn't going to make running your node 2 times faster. Things don't scale this way, and the most important thing you need to remember is that aside from central exchanges, some wallets, and mining pools, everyone else is running their node at a net loss. So why would you want to make life more difficult for those maintaining the ecosystem just so that the Orindals' fools get to protect their useless shit from pruning?"

[larry_vw_1955]

I have said that I, and all of us, would rather leave Ordinals as they are, than moving them to the UTXO set. I don't think anyone would disagree with this statement. Ordinal users would continue clogging up the network with trash, it's just that this time, we'd have an extra burden to account for; an inflationary UTXO set. Why should we incentivize ourselves towards that direction?

i understand your viewpoint of letting ordinals continue to exist so that people don't try and use something like bitcoin stamps more but if bitcoin is robust, then it shouldn't matter how many people want to use bitcoin for multisignature transactions using bitcoin stamps or any other multisignature software.

If something can't be judged as good or preferred according to you, then why are you arguing that Ordinals should be moved to the UTXO set?

ordinals can't be moved to the utxo set. that is not how they were designed. the true test though would be if ordinals ceased to exist and then these people would have to store their monkeys and things in real transactions. multisignature or whatnot. and be paying real fees. that would take care of itself most likely. as it always has. but you don't want to put things to the test you would rather allow the lesser of two evils to exist so that bitcoin could never be put to the ultimate test. ok. 

A large multisig, done with OP_CHECKMULTISIG is harmful, because it has at least O(n^2) complexity (it can be bigger, if you have a lot of multisig transactions, connected with each other).

so what are you saying exactly, that you wish OP_CHECKMULTISIG was not a part of bitcoin? you do realize it is not going anywhere but i think taproot multisignature is cheaper in terms of fees so people might opt for it instead. to your approval.

You're missing the point. You asked why a large UTXO set is a problem, and your solution is to buy more RAM? So in the same post in which you argued Linux is removing support for systems with 4 MB RAM, you argue 8 GB isn't enough. I'd argue adding more RAM doesn't scale well: what if there are 100 times more UTXOs? Add 2 TB RAM to run Bitcoin Core?

I think 16GB of RAM is plenty right now. In 3 years, you upgrade to 32. 3 more years, upgrade to 64. Does that not sound reasonable because it does to me. Applications are going to increase in their system requirements there's no doubt about that. Thinking you can really get by with 8GB forever is not really realistic in my opinion.

Downloading isn't resource intensive, verifying all data is what gets you.

yeah that doesn't sound too fun. having to download other peoples' monkeys and not only do that but also verify them too. just so that you can prune the monkeys later on.

It won't be faster, and it won't reduce system load. Besides, if you downloaded one wrong bit in a block, you'll have to start over.

sounds like a nightmare having to run bitcoin core. to be quite honest.

[DooMAD]

Downloading isn't resource intensive, verifying all data is what gets you.

yeah that doesn't sound too fun. having to download other peoples' monkeys and not only do that but also verify them too. just so that you can prune the monkeys later on.

But it's the price you pay to remove the need for trust.  You wouldn't have to do any of it if you were happy to rely on people being honest, but that wouldn't be the trustless and secure network we've built.  We know it's inefficient, but it is vital. 

Plus, some people are offering to do all that for you so that you don't have to.  Maybe show some appreciation by not asking to make their task any more challenging than it already is.   

It won't be faster, and it won't reduce system load. Besides, if you downloaded one wrong bit in a block, you'll have to start over.

sounds like a nightmare having to run bitcoin core. to be quite honest.

There are merely practical considerations people need to be aware of.  Idealism has its place, but it works like it does for good reasons and I get the sense that people don't fully appreciate those practical considerations when they ask for changes.

[LoyceV]

If you happen to own computer with big RAM which can store all UTXO, using HDD should be fine.

I can confirm this, 32 GB RAM and HDD works just fine

Another point worth mentioning is that spending 2x on your RAM or anything else isn't going to make running your node 2 times faster. Things don't scale this way

If you have enough RAM, you're right. But if your system is short on RAM, doubling the RAM could improve the performance a lot more than just doubling it.

I think 16GB of RAM is plenty right now.

I'd like to have more, but my laptop can't handle it.

In 3 years, you upgrade to 32. 3 more years, upgrade to 64.

Many laptops are still sold with 8 GB nowadays, and from what I've seen, memory stopped following Moore's law in the past 10 years. It's no longer doubling every 18 months, it now takes a lot longer to see "standard" new laptops double in memory.

Does that not sound reasonable because it does to me. Applications are going to increase in their system requirements there's no doubt about that. Thinking you can really get by with 8GB forever is not really realistic in my opinion.

I never claimed 8 GB is enough forever. I'm saying it's what many people use (and many others use less than that). I haven't checked the hard data, but I think the UTXO set has increased several gigabytes in just the past year. That's not scaling well with RAM-upgrades.

It won't be faster, and it won't reduce system load. Besides, if you downloaded one wrong bit in a block, you'll have to start over.

sounds like a nightmare having to run bitcoin core. to be quite honest.

Maybe you should try it. There's a reason Bitcoin Core downloads and verifies blocks at the same time, and it works quite well.

[larry_vw_1955]

Many laptops are still sold with 8 GB nowadays, and from what I've seen, memory stopped following Moore's law in the past 10 years. It's no longer doubling every 18 months, it now takes a lot longer to see "standard" new laptops double in memory.

i'm sure you can buy desktops with even more then 128GB of RAM. but this one can be configured with 128.

https://www.dell.com/en-us/shop/desktop-computers/optiplex-small-form-factor/spd/optiplex-7020-plus-small-ff/gctoo7020sffpus_vp?redirectto=SOC&configurationid=9e300b86-379e-4ddb-8b2c-d9e498e02ccc

there will always be a market for lower end stuff that only has 8GB. that's for people who just want to send some emails and watch a few youtube videos. nothing more...and there's plenty of places that will sell you something like that just realize, you won't be able to do much with it. not as time goes on.

But it's the price you pay to remove the need for trust.  You wouldn't have to do any of it if you were happy to rely on people being honest, but that wouldn't be the trustless and secure network we've built.  We know it's inefficient, but it is vital. 

well yeah i mean bitcoin is definitely inefficient. it uses alot of electricity worldwide. but i dont know of a more efficient way of having a decentralized payment system.

Plus, some people are offering to do all that for you so that you don't have to.  Maybe show some appreciation by not asking to make their task any more challenging than it already is.   

bitcoin sure is lucky so many people are willing to store the blockchain FOR FREE. and not only store it but let other people use their internet connection to download things from them. and let them send their transactions to them for forwarding to other people. sounds like slavery to me. 

There are merely practical considerations people need to be aware of.  Idealism has its place, but it works like it does for good reasons and I get the sense that people don't fully appreciate those practical considerations when they ask for changes.

customers are always going to be asking for new services and things. and no they might not always understand underlying issues. like people working for no pay. but that's a problem with bitcoin and how it was designed. i'm not sure we have the right to tell people "no, you can't do that using bitcoin. because i'm running a node and no one is paying me and i don't want to store even more data." customers would not understand why someone would work for no pay. and donate their time and resources. and that's the beginning of being treated with no respect, like a doormat. for people running full nodes...

[DooMAD]

But it's the price you pay to remove the need for trust.  You wouldn't have to do any of it if you were happy to rely on people being honest, but that wouldn't be the trustless and secure network we've built.  We know it's inefficient, but it is vital.  

well yeah i mean bitcoin is definitely inefficient. it uses alot of electricity worldwide. but i dont know of a more efficient way of having a decentralized payment system.

But in this instance we're not talking about electricity, but the tradeoff between efficiency of verification and not needing to rely on trust.  That's one of the key pillars it's all built upon.  Something we must never lose sight of.  We accept inefficiency because it means we don't have to trust.  

bitcoin sure is lucky so many people are willing to store the blockchain FOR FREE. and not only store it but let other people use their internet connection to download things from them. and let them send their transactions to them for forwarding to other people. sounds like slavery to me.  

It can't be slavery if people are offering to do it and have the option to stop at any time.  And people aren't doing it because they expect a monetary reward.  It's again about not needing to rely on trust.  The people running nodes benefit from that.  It's their primary incentive.  SPV users don't gain that benefit.

[vjudeu]

SPV users don't gain that benefit.

Exactly.

1. Usually, full nodes can see new blocks faster, than they appear on block explorers (because other nodes have a reason to connect to them, and to tell them that information before, and get some other data in exchange). For example: you can see some block in your node, and then, 15 or 30 seconds later, it will appear on mempool.space.
2. Full nodes receive and send transactions all the time. Which means, the whole traffic is not related only to their own transactions, but to the whole flow of new transactions in mempools.
3. People can connect and disconnect at will. As well as switch their mode into pruned, archival, or change their settings. For example: you can allow any fee, and then see the traffic below 1 sat/vB.
4. Full nodes can observe all kind of weird behaviour, related to the network, for example violating sigops limit: https://bitcointalk.org/index.php?topic=5447129
5. If you have a full archival node, then you can observe the chain, without telling block explorers, which transactions are you interested in (and which of them are potentially related to your own payments).

So, it is not a slavery. But: if you verify the whole chain once, then you have all of that data. And there are incentives, to run just some pruned node (and not pay for example for renting additional disks in VPS), instead of running a full, archival node. And that makes it harder to create new nodes, because then you have less peers to get blocks from.

So, if people will abuse the chain by pushing a lot of data, not related to transactions, or even deploy some altcoins on Bitcoin, or do some other weird things, then guess what: some full node operators can decide to just use pruning, and not to care about new nodes. What then? How people would access their Ordinals, if there would be not enough nodes to fetch that data from?

And the same is true for the UTXO set: if it will be abused, then running a node may require also pruning some parts of the UTXO set, and replacing it with some weaker proofs. What then?

[BlackHatCoiner]

i understand your viewpoint of letting ordinals continue to exist so that people don't try and use something like bitcoin stamps more but if bitcoin is robust, then it shouldn't matter how many people want to use bitcoin for multisignature transactions using bitcoin stamps or any other multisignature software.

But... It doesn't matter already. What makes you think it does matter?

ordinals can't be moved to the utxo set. that is not how they were designed.

Ordinals can be moved to the UTXO set. Take an Ordinal, split it in an array of 256 bits, convert these bits to addresses respectively, send 0 sat to each address. You just created another representation of the same ordinal.

bitcoin sure is lucky so many people are willing to store the blockchain FOR FREE. and not only store it but let other people use their internet connection to download things from them. and let them send their transactions to them for forwarding to other people. sounds like slavery to me. 

Storing and sending/receiving blocks and transactions are not done "for free". This is the only way to be certain that your ledger is intact, pure, true. And it's not that it costs a thousand dollars to do it. You just need a decent Internet connection, any modern computer and a terabyte.

[cryptosize]

I guess torrent seeding is considered "slavery" too, since there's no monetary benefit...

People forget that back in 2009 nodes & miners were the same thing (due to CPU mining), even though they are decoupled today, because ASICs are quite expensive.

[larry_vw_1955]

i understand your viewpoint of letting ordinals continue to exist so that people don't try and use something like bitcoin stamps more but if bitcoin is robust, then it shouldn't matter how many people want to use bitcoin for multisignature transactions using bitcoin stamps or any other multisignature software.

But... It doesn't matter already. What makes you think it does matter?

because you merited this posting. so took that to mean you don't agree with bitcoin multisig being used as the underlying mechanism for bitcoin stamps. if there is some issue about bitcoin multisig then that's the developers fault not people trying to use the feature however they want to. that's kind of my opinion.

a built in feature of bitcoin (multisig) being used in way that you think might be harmful

It is harmful. A large multisig, done with OP_CHECKMULTISIG is harmful, because it has at least O(n^2) complexity (it can be bigger, if you have a lot of multisig transactions, connected with each other)....

Storing and sending/receiving blocks and transactions are not done "for free". This is the only way to be certain that your ledger is intact, pure, true. And it's not that it costs a thousand dollars to do it. You just need a decent Internet connection, any modern computer and a terabyte.

i don't know. to me it just seems like the people that are supporting the network by running full nodes and letting other people connect to them and download blocks or relay transactions, they should be getting a piece of the pie. not just the miners. you can't just tack on a feature like that later on though. it would have had to have been designed that way from the beginning.  

I guess torrent seeding is considered "slavery" too, since there's no monetary benefit...

which is why once i downloaded some torrent, i immediately shut down the torrent software because i got what i wanted. why am i going to sit there being a server to other people downloading it? i know that's a bad attitude to have but that's the way alot of people are. probably why most torrents eventually dry up and no seeders are left. that could happen to bitcoin too possibly unless people see some kind of benefit from serving up the blockchain to other people for free...

[cryptosize]

Storing and sending/receiving blocks and transactions are not done "for free". This is the only way to be certain that your ledger is intact, pure, true. And it's not that it costs a thousand dollars to do it. You just need a decent Internet connection, any modern computer and a terabyte.

i don't know. to me it just seems like the people that are supporting the network by running full nodes and letting other people connect to them and download blocks or relay transactions, they should be getting a piece of the pie. not just the miners. you can't just tack on a feature like that later on though. it would have had to have been designed that way from the beginning.  

I guess torrent seeding is considered "slavery" too, since there's no monetary benefit...

which is why once i downloaded some torrent, i immediately shut down the torrent software because i got what i wanted. why am i going to sit there being a server to other people downloading it? i know that's a bad attitude to have but that's the way alot of people are. probably why most torrents eventually dry up and no seeders are left. that could happen to bitcoin too possibly unless people see some kind of benefit from serving up the blockchain to other people for free...

You're the epitome of FYIGM.

When you hodl BTC, it's in your best interest to not let it die and give something back (even if it's just a node/blockchain explorer).

ps: That's why the best torrent trackers are private and strictly enforce a 1:1 ratio. So yes, you can tack on such a feature later on, although it would have been much better if game theory was built-in (like PoW does).

For me it's not a big deal... if you generally leave your computer 24/7/365 online, it doesn't hurt to seed a bit more. It's not like I consume more electricity (that's why ASICs are a deal-breaker for regular home usage, unlike RPi nodes).

[LoyceV]

And the same is true for the UTXO set: if it will be abused, then running a node may require also pruning some parts of the UTXO set, and replacing it with some weaker proofs. What then?

Then you can no longer validate a block that includes one of the pruned UTXOs. That's why I don't think UTXO pruning will ever be a thing. If they're unlikely to be used, they don't need to be in RAM. But in the unlikely case you need one anyway, it's nice if you still have it on disk.
For normal nodes it's 2% of the total blockchain. But I can see a problem for pruned nodes: the UTXO ("chainstate") is already 20 times larger than the blocks stored in the smallest possible pruned node.

[BlackHatCoiner]

because you merited this posting.

OK, I see. It's still a little unclear, though. If a transaction takes a long time to verify, then it can act as an attack vector. Disabling it is more of an argument in favor of robustness. Multi-sig transactions are allowed, but with M, N limits. For example, AFAIK, you can't spend from a 1000-of-1000 multi-sig.

they should be getting a piece of the pie. not just the miners.

The work is done by miners, and the pie is split on that work. If you think running a node helps on bandwidth and needs a piece of the pie, then what happens if you spin up 10,000 nodes?

[LoyceV]

If you think running a node helps on bandwidth and needs a piece of the pie, then what happens if you spin up 10,000 nodes?

We'd get millions of virtual nodes created by the same person to get a larger piece of the pie, and nodes will become as centralized as mining pools are now.

[ABCbits]

I guess torrent seeding is considered "slavery" too, since there's no monetary benefit...

which is why once i downloaded some torrent, i immediately shut down the torrent software because i got what i wanted. why am i going to sit there being a server to other people downloading it? i know that's a bad attitude to have but that's the way alot of people are. probably why most torrents eventually dry up and no seeders are left. that could happen to bitcoin too possibly unless people see some kind of benefit from serving up the blockchain to other people for free...

Mining pool, block explorer, some exchange and service (which offer API to obtain data from blockchain) need to run full node to keep their business running. So scenario you mention is unlikely to happen, unless Bitcoin itself no longer attractive or useful.

If you think running a node helps on bandwidth and needs a piece of the pie, then what happens if you spin up 10,000 nodes?

We'd get millions of virtual nodes created by the same person to get a larger piece of the pie, and nodes will become as centralized as mining pools are now.

Or worse, one or few full node that use multiple IP address/port. See https://reddit.com/r/bitcoinxt/comments/3iao3i/how_to_run_3000_completely_legit_full_nodes_aka/.

[larry_vw_1955]

because you merited this posting.

OK, I see. It's still a little unclear, though. If a transaction takes a long time to verify, then it can act as an attack vector. Disabling it is more of an argument in favor of robustness. Multi-sig transactions are allowed, but with M, N limits. For example, AFAIK, you can't spend from a 1000-of-1000 multi-sig.

well yeah obviously, there's got to be some limits but no one is doing larger than M=N=15 most likely and i doubt people even do anything that big. But there might be 5 of 7 but i think bitcoin stamps only does 2 of 3. so there's not going to be any justification for saying that 2 of 3 multisig transactions even the ones from bitcoin stamps are not welcome on the bitcoin network.  

The work is done by miners, and the pie is split on that work. If you think running a node helps on bandwidth and needs a piece of the pie, then what happens if you spin up 10,000 nodes?

if actual work is required of a node to validate and process a transaction in some way then i don't think it would be feasible to spin up 10,000 nodes.

We'd get millions of virtual nodes created by the same person to get a larger piece of the pie, and nodes will become as centralized as mining pools are now.

for example, say you changed bitcoin transaction format so that there had to be an extra second signature on top of the first signature. the second signature would be from a node that validated the first signature. the node that did the validation would get a small reward if the transaction they signed and validated made it into a block.

if a node didn't want to participate in the signing process they could just broadcast it out to other nodes and let them sign it. so it could be like an opt-in system for nodes who are looking to pick up a small amount of spare cash. 

[LoyceV]

for example, say you changed bitcoin transaction format so that there had to be an extra second signature on top of the first signature. the second signature would be from a node that validated the first signature. the node that did the validation would get a small reward if the transaction they signed and validated made it into a block.

This makes no sense. Every node has to validate every transaction. If a node earns money from signing, miners will sign everything by themselves and take the money.

[vjudeu]

if actual work is required of a node to validate and process a transaction in some way then i don't think it would be feasible to spin up 10,000 nodes.

It depends, which software you use, and how do you configure it. If you spin 10k independent nodes, each in a separate directory, using Bitcoin Core, then it is hard. But if you only download everything once or twice, and everything else will be related just to some proxies, used to redirect the traffic properly, then it requires not much more effort, than running a single node. In that case, you can have one or two full nodes, and a lot of SPV-like nodes, which are connected into that, and just create a lot of traffic out of thin air, if you can allocate many IPs.

Because note one important thing: nodes validate data, because then they know, if something is correct or not. But if your one or two nodes already know, that transaction X is correct, then 9998 other nodes don't have to repeat that computation. They are all owned by a single entity, so they can trust each other.

the node that did the validation would get a small reward if the transaction they signed and validated made it into a block

1. This is called Proof of Stake, and has its own problems.
2. You have to find a way to distribute that "small reward" efficiently, without creating additional outputs (just like transaction fees are distributed).

if a node didn't want to participate in the signing process they could just broadcast it out to other nodes and let them sign it

There are bigger problems: a single input would then have thousands of signatures, unless you will join them somehow. Or: a single node can easily pretend, that 10 nodes validated something, because there are valid signatures for 10 different public keys.