NFTs are only spec for smart contracts that dictate mappings between "owners" and "token ids". That's it. End of story. There isn't even a universal requirement

NFTs are only spec for smart contracts that dictate mappings between "owners" and "token ids". That's it. End of story. There isn't even a universal requirement that maps from token id to metadata like OpenSea's format. > NFTs can be minted in exchange for currency and can be traded between owners later The part of dApps that recognizes NFT ownership for releasing funds is just business logic that can be replaced with anything we can imagine - including but not limited to checking the storage for an Address->RandomString lookup (in fact, if you squint, that _is_ what's happening in transactions due to the step that verifies wallet signatures via cryptographic hashes) There is nothing at all in the spec that allows for them to have any special meaning in terms of vouchers for currency > More complex rebuttal: you can get peer-to-peer (and even peer-to-pool) loans using your NFT as collateral I don't think you understand how NFTs work. This isn't even a more complex rebuttal, NFT contracts don't care if the owner is another contract or a human with a wallet. They only care about the owner to token id mapping. If anything, p2p loans is a _simpler_ example because you're moving the logic out of contracts and into, perhaps, humans who decide to act on nft transfers. It's a weird example imho, but regardless, it's only an extension of the above points. > How is any of this applicable right now to a HTTP header?! In HTTP, you acquire a random string (via some web app), and then you later provide it to prove authorization In Blockchain, you acquire a random public/private keypair (via wallet generation), and you later sign data with this random key in order to prove authorization In NFTs, you use the Blockchain infrastructure to sign transactions that associate you as the owner with arbitrary token ids. Do you not see that it's fundamentally the same thing? > Also, the only people I ever heard of mentioning “videogames fully on chain” are crypto opponents; they try to come up with the most bizarre scenarios, to “prove” that crypto is inadequate. It’s so irrational. Ok. I mean, I've hands-on deployed CW721 contracts on cosmos with a Unity frontend for on-chain gaming demos. Your ad-hominem attacks are pretty weird from where I'm sitting and, frankly, reek of "dunning-kruger effect". But I don't want to get into a pissing contest, let's stick to facts: 1. Most chains require you to sign transactions all the time. It's a crappy UX. Near avoids this to some extent, and it's a clientside thing that can be skipped, but generally at the expense of security most users aren't comfortable with. 2. Transaction time is slow af compared to what you get with regular socket servers. I'm talking a difference of speed in terms of decades of progress. Not years. Decades. 3. Nodes go out of sync, you have race conditions, there are real problems to solve that robust game server tech deals with and blockchain doesn't concern itself with. Working around these are doable but painful and not out of the box. 4. Most games depend on randomization and/or hidden state that clients cannot be made aware of to prevent cheating. There are some chains like Secret Network that propose solutions to this, but overall, blockchain is transparent in ways that are showstoppers for on-chain games. > Yeah, it’s almost like blockchains are designed to be financial infrastructure. At least you got the basics right, I guess. Yup.