Though I guess you’re also using the op_return tx to reduce spam.. could a pow on the notes help solve that instead of all the onchain tx?
Discussion
It’s not just to reduce spam, though that helps. Bitcoin provides absolute ordering. Outputs inside transactions inside blocks, all in exact sequence. This allows indexers to verify who claimed a name. POW doesn’t provide any kind of ordering guarantee
If two notes were confirmed in the same block for the same new id via OTS then PoW used to reduce spam maybe the same PoW nonce could also be used to break ties? But i do see how using position in a block as being simpler.
It is simpler which is a big win, but the main reason what you said wouldn’t work is because someone could claim a zillion names at no cost and never publish a nostr event to prove it. Then when someone decided to claim a name that the spammer had claimed (he could fit a million in the OTS transaction if he wanted), he could just reveal his first “claim” and rugpull the whole protocol.
In Nomen, there is a 5 byte fingerprint in the metadata. The fingerprint is the first five bytes of the HASH-160 of the name. It’s right there, in the open, on the blockchain. You can tell if anyone has claimed a name you want without publishing a Nostr event just by scanning the blockchain for your fingerprint. No chance of getting rug pulled by a sneaky squatter.
I don’t think footprint is a huge problem with this protocol. On Nomen output is 41 bytes in total. That means if every single person on earth registered one, it would only add 370 GB to the blockchain. It’s not like you need a lot of these names, or need to make frequent transactions on chain. Of course, that isn’t possible, anyway. Long before that happened, a name will be too expensive to register on chain for almost anyone, and they will move to sidechains (I talk about this at the end of the post).
This is a good feature! I look forward to seeing how it develops. Anything that could replace centralized authorities like ICANN is worth pursuing.
I think I just maybe understood what you mean. Yes, I am aware of opentimestamps. What makes it a non-starter is that it is compresses infinite documents into a single hash (as I understand it). This is not desirable because then someone could claim many names at no cost. Which is probably closer to what you meant, I am guessing? If so, yes, that is why I chose the method I did
Yes this is what I was thinking and why I suggested pow on the notes could be used instead for spam reduction. Motivation being to reduce onchain footprint except minimal needed for ordering notes.