Topic: NFTs in the Bitcoin blockchain - Ordinal Theory - page 31. (Read 9515 times)

But that's not actual demand, that is pretty much like the same thing with a new name or disguise. If you are saying that the garbage being created today is going to lead to more garbage makers coming to bitcoin, then that is not demand and it is exactly the points everyone is raising against Ordinals.

I believe, from their viewpoint/perspective, there were also other NFT that brought demand, like CryptoPunks or Bored Ape Yacht Club. To play Devil's Advocate, but this is also a fact, the early NFT brought a new kind of market that brought demand for the blockchains they are in.

It could be used as an attack vector on the network, but if you're debating that it will not bring demand, then that's probably going to be proven wrong, it will be a problem.

Thanks! Interesting discussion. Seems that many main developers agreed with the limit-removing idea.

This is an interesting point. OP_RETURN-style inscriptions allow the ownership being transferred entirely without using UTXOs at all for this purpose, because they can use a separate database to store ownership information (and other stuff, of course). Obviously most "ownership-transferring" txes would still create UTXOs but only due to the fact that you need an BTC input for the fees, so they won't probably be cluttering the UTXO set for a long time and be spent eventually in a "financial" transaction.

In the Ordinals implementation you will have always one UTXO "residing" at the current NFT owner, and NFT ownership may be commonly longer than ownership of "normal" fungible Bitcoins, so the cluttering effect will be probably higher.

This would be in fact another argument in favour of making OP_RETURN cheaper than the Ordinals stuff. AFAIK the main problem to achieve this that OP_RETURN has no witness discount, so if this is not changed, Taproot/Ordinal inscriptions will be still cheaper from a certain data limit upwards.

Cryptokitties and similar garbage didn't bring any kind of real demand for the platforms they were on. I don't see how Ordinals is going to be any different! Not to mention that we already had better ways of creating tokens in sidechains and so far a handful of people even talked about them and the only reason why Ordinals is being discussed is because it is spamming the chain and is a threat to the real demand bitcoin has.

That's merely signalling for Ordinals/Inscribing/NFTs in the Bitcoin Blockchain because "they" probably think that long term, the success of Ordinals might bring higher demand to use Bitcoin, then therefore higher profit for mining, which will definitely increase the total hashing power, and then also increase the network's security.

The main utility for Ordinals for those haters of Bitcoin Core, and the Bitcoin Core Developers, is they can inscribe dick pics and know that they will be in gmaxwell's node, achow's node, everyone's nodes.

 

That's what developers have been discussing recently on the bitcoin-dev mailing list [1] and in fact they seem to agree that the best course is to remove the OP_RETURN limit (which are part of the standard-ness test, not a consensus tule) altogether [2].

[1] https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2023-February/021387.html

[2] https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2023-February/021438.html

I'm still reading stuff about it to really be able to form my opinion on the Taproot "inscription" thing (Ordinals itself seem not to have so much to do with the controversy, they're simply an alternative and efficient method to associate a number/name to the digital object/NFT which is stored).

For now I'm not too concerned, but I have some questions maybe someone can answer or point me to links where they're been explained.

The first big question is of course: how much fees do people save with this method, compared to traditional methods (OP_RETURN with multiple transactions, and Segwit inscriptions like Peter Todd's publish-text.py)?

This tweet claims the reduction compared with the "Segwit method" is 15%. Can anybody confirm this? I also think that probably the amount saved could increase with the size of the digital object, is this true?

Second question: I am a bit concerned that these Taproot scripts cannot be pruned (although from my understanding they aren't cluttering the UTXO set as they are published/stored only after the transaction has been spent), while OP_RETURN scripts can. (if wrong maybe someone can point me to the correct understanding).

So, from my perspective, the way to "solve"/"reduce" the possible spam problem would be to try to ensure OP_RETURN to become the cheapest on-chain method to store data again. Maybe the OP_RETURN standard limits could be increased, while the Taproot limits can be decreased at the same time?

---

PS: Just my personal opinion - Namecoin is much better fitted for the whole NFT stuff. Why don't the Ordinals folks move there? NMC has always been about names and identities, it's blockchain is quite empty and it's even merged mined with BTC.

I agree that being non-standard or not isn't the way of stopping spam, but I strongly disagree with messing around with op codes, hardening consensus rules which locks out potentially useful functionalities to get rid of this particular batch of NFT mania, which will overtime find their way in otherwise.

Bitcoin is incentive based, and pro-freedom network. If transaction fee can't discourage a spammer, no thing can.

Pre-taproot, ordinal NFTs could also have used segwit for similar cost. Being non-standard might discourage them, but cannot stop them as long as they can be submitted directly to a mining pool willing to include them, as is happening with ordinals.
And even pre-segwit, ordinals could have deployed at a 4x higher cost (without the witness discount).

So you're right about being able to discourage them, but wrong about being able to stop them.

Very unlikely to happen when most node owner don't bother change default setting. Additionally, in past there were proposal to change default relay fee to 0.1 sat/vB for Bitcoin Core.


Generally it's true, although we have to consider few people willing to pay more since Bitcoin is one of most popular cryptocurrency.

We have been stopping them successfully before Taproot by making them non-standard which discourages such spam. Otherwise you could create a scriptpub with an arbitrary input that is a little smaller than 1 MB.

A Schnorr signature is 64 bytes and with the public key it becomes 96 bytes. 80 byte limit is not enough for any type of Taproot witness. You also can't use OP_CHECKMULTISIG in Taproot scripts, OP_CHEKSIGADD is used so the 15-of-15 size is not the same as before needing all 15 keys and signatures.

You are wrong. A lot of forward compatible code existed in bitcoin from day one. That is how BIP16 was enabled too. It allows us to remain backward compatible in each soft fork.
Another example is the the simplest OP codes called OP_NOPs that were used in 2 soft-forks to enable OP_CHECKLOCKTIMEVERIFY (BIP65) and OP_CHECKSEQUENCEVERIFY (BIP112).

It can be non-standard, but it cannot be invalid, because it could invalidate currently existing timelocked transactions with too huge witness. Also note that you can use Schnorr signatures and have a single signature only if your sighashes are the same: so if you have 2-of-2 multisig, where you have SIGHASH_ALL in one signature, and SIGHASH_SINGLE|SIGHASH_ANYONECANPAY in another signature, then your proposal will block those use cases, because then you need a separate signature for each kind of sighash.

Every single spam transaction can be splitted into multiple spam transactions. Note that people can still use zero satoshi outputs, and put many 80-byte chunks, to generate the same spam, and take even more space, because of transaction format. There are more ways to create spam than you can count, so it is impossible to prevent all of them. If something makes it harder for typical multisig users, and does nothing to stop spam, then it should not be introduced, because the final outcome will be worse than what we have today.

We already have that check in our standardness rules. Note that this huge NFT transaction is non-standard, because of many reasons. One is non-standard transaction size, another is non-standard transaction fee, and so on.

There is a pretty straight forward solution to this, albeit a somewhat brutish one:

The Ordinal protocol sets out to identify single satoshis. It therefor generates outputs that are provably self-unspendable: their value isn't enough to cover for the fee to spend them. To redeem a 1 satoshi output, you would have to create a 0 fee tx, or bring in coins from another output.

If a majority of nodes reject these transactions, this whole ordinal idiocy goes tits up overnight. This can be done without any changes to consensus rules: set your node to a reasonably s/B fee rate. This would limit the propagation of these transactions in the mempool.

A stronger iteration of this approach would be a UASF where nodes would reject transactions that create such outputs. If 2/3rd of the nodes were to enforce such rule, this would force miners to upgrade their nodes as well, as they would rather stay on the main chain than mine ape jpegs. Then the damage can be undone by evicting these outputs from the mempool with a subsequent update.

In general, I'm not too worried. Ethereum has demonstrated that NFTs will move to the next chain down the road when fees get too high for data usage. In general the whole concept of NFTs on chain is dumb, as you need a centralized party to enforce consumption rules. I don't see this fad developing on chains with scarce block size as a consequence.

Oh noes, please keep that NFT nonsense off Bitcoin network. Puuuuleeeeeeaaaz. I think Ethereum is a perfect blockchain for that NFT crap and they should stay there. 

I must admit I've read this handbook but I still don't understand why Bitcoin and bitcoiners would need these "ordinals". Humans are collectors is a very vague description. I wouldn't pay 0.5btc for a satoshi, just because somebody called this satoshi "satoshi" or assigned a number to this sat. I'm sorry but that's stupid.

I'm all about that too. I want from the overwhelming majority of the Bitcoin users to compress their transactions best. I neither want them to store trash, but I want that less than I want Bitcoin to be a completely pro-freedom and censorship resistant protocol.

And since the only manner to disincentivize storing arbitrary data on-chain without discarding these principles is the transaction fee, I like better leave it there. In the end, a fool and his money are soon parted.

As I said before, I'm all about maximizing the amount of monetary transactions while keeping blockchain size low, since that's directly correlated with decentralization staying high.
I know that right now, if you want to spam the blockchain, as long as you pay for it, you can. But I personally believe it is desirable that the extremely limited blockchain space (total blockchain from 2009 to 2023 is smaller than any one of all of my personal HDDs and SSDs) is used in the most efficient way possible.
That is my personal opinion, of course. I can't force it on anyone.

The title mentions electronic cash specifically. There is no other such uncensorable p2p electronic cash system with the same amount of decentralization and security as Bitcoin, so it would be foolish 'sacrificing' it for NFTs (even though they may be ready to pay a lot of money).

Since Luxor mint NFT with 3.94 MB on their block which has Ordinal ID/number 652[1], now current highest Ordinal ID/number is 6681[2]. While not all of them have transaction size >300KB or ~3.9MB, it's clear 1 sat/vB isn't enough to discourage people to mint BTC on main layer.


Why 80 bytes? I expect some multi-signature address which use P2SH or P2WSH would unable to spend their Bitcoin if this limit is enforced on protocol level.

---

[1] https://ordinals.com/inscription/0301e0480b374b32851a9462db29dc19fe830a7f7d7a88b81612b9d42099c0aei0
[2] https://ordinals.com/inscription/d5c8c24efbbea3239acb93dfe1b0cc45dc419b76cd22690bfb5aff43abec34cci0

The ordinals experiment has clearly shown that you can't stop non-transfer uses within the existing consensus rules. As long as there's script, and as long as there are addresses, then there are straightforward ways to encode arbitrary data within them. Only a script-free blockchain without adresses, such as Mimblewimble, can strongly limit the amount of arbitrary data, to about 2% of chainsize (using lock heights in the past).

And expensive fees, which would discourage them to use it as currency. Again, what's the problem? The fact that someone decides to play by the rules and pay for stuff you don't approve of, or that the Bitcoin network can't handle lots of on-chain transactions cheaply?

I don't believe the rules we have (besides the consensus) limit an activity with high demand. As said by pooya87, standard rules are soft rules. If someone desperately needs to make a non-standard transaction, he will. And theoretically, miners should allow any transaction that's in favor of their pocket, which means to follow minimum to no standard rule.

It is more accurate to say it can become too expensive otherwise the 1 sat/vbyte fee that is paid most of the times is not expensive at all and that's the problem. They can abuse this very cheap cost (cheaper than ETH fees) to spam and increase it.