Is the artwork actually stored in the token or does it contain a URL that isn't guaranteed to always exist? I'm guessing there is no mechanism to change the link if the host should go down or stop hosting the content.
Are NFT's used to recognise a transfer of ownership of the intelectual property?
* The artwork is just pointed to, it's not on the blockchain.
* The URL may or may not exist for any given length of time either. (Even some NFTs that tried to point to IPFS, actually pointed to an IPFS redirector.)
* No other rights - copyright, moral rights, reuse rights, etc - are conveyed without an explicit contractual transfer. Even the Christie's deal says, once you dig through the 33-page sales agreement, that you are just buying the token itself, and not the image pointed to.
Also, you have zero guarantee that the artist had anything to do with that particular NFT - and there's a lot of NFT grifters "minting" other people's art.
The metadata in the NFT can (and in my opinion _should_) point to an ipfs:// style URL. The websites displaying the NFT would have to use some redirector, but the actual token would have a URL that anyone could host (known as "pinning") on the IPFS network.
Remember that IPFS is functionally just BitTorrent with magnet: links. So if nobody's seeding the file ... it's not there any more. And it looks like a pile of literally the sites selling NFTs don't bother seeding the files they've sold, mere weeks later. Thread: https://twitter.com/jonty/status/1372169695277760519
Very few current NFTs exist on-chain. A handful do, and there's the interesting case of generative art where the code that makes the art lives on-chain.
There's a push to get things on Arweave, a solution that purports to be a perpetual storage solution. In the long run, I think all art NFTs will have to go in that direction to remain credible.
Are you saying the underlying asset isn't stored in the blockchain? Or that the "title" isn't stored in the blockchain?
How is the title associated with the asset? Presumably the title is something like a sha256 hash associated with a wallet. How do you know that hash means "beeple artwork"?
Christie's interface didn't just lead with the transaction hashes, but your conclusion is pretty far off, for now.
Try emailing them next time.
I'm understanding of the frustration, it took me a year of asking similar questions about Venezuala's Petro coin, which was even subject to pre-emptive US sanctions, before it became official that the Petro didn't exist and was never issued. Nobody could tell me the contract address and people acted like I grew two-heads for even asking, when it should have taken two seconds.
The technology allows for really basic things to be transparent, and when you don't lead with that stuff and nobody asks about that stuff, it does become pretty obvious that they aren't even using the technology.
Enter 40913 in tokenURI at https://etherscan.io/token/0x2a46f2ffd99e19a89476e2f62270e0a...
You will see an IPFS URI which contains the metafile.json: https://cloudflare-ipfs.com/ipfs/QmPAg1mjxcEQPPtqsLoEcauVeda...
This in turn contains a URL to the image: https://ipfsgateway.makersplace.com/ipfs/QmZ15eQX8FPjfrtdX3Q...
https://www.christies.com/img/LotImages/2021/NYR/2021_NYR_20...
Whereas the IPFS version of the image is merely 1200x1200. In this case, it really matters because the image is a mosaic that is virtually indecipherable at the 1200x1200 size.
So, it begs the question what is the person actually buying? I get that it's the concept of "ownership" that's being sold, but ownership of what exactly?
When the new owner then goes to sell "it", what do they reference? An illegible JPG?
The NFT doesn't give you jack shit. As the Conditions of Sale [0] says:
> You acknowledge that ownership of an NFT carries no rights, express or implied, other than property rights for the lot (specifically, digital artwork tokenized by the NFT).
[0] https://www.christies.com/pdf/onlineonly/ECOMMERCE%20CONDITI...
Yeah, interesting that this notion of ownership is separate from IP. And, of course anyone can download the file itself.
So, they're essentially selling the right to declare you own it and the ability to sell that right to someone else.
tho, the important piece is the network. the network doesn't know or care if random people have a locally downloaded copy of the art. all sites, services, validations, virtual worlds, whatever on the network will hit the ethereum contract which gives up an address which you've proved you own by key signature.
the network can't see nor does it care about your local files.
which means, basically, any other appearance of the art in any other place is just an ad for the current owner of the nft.
What even is the "asset" in a lot of cases?
People are buying Tweets, which are generated in real time by querying a database and generating some HTML and CSS. Or some UI code on native platforms. Or some JSON via an API. What is the asset? The Twitter code that generates the Tweet? A JPEG of the Tweet itself? The text of the Tweet?
If it's an image of the Tweet, does take into account whether or not my user agent is in light or dark mode?
The rabbit hole is never-ending with such questions.
(enter 40913 at https://etherscan.io/token/0x2a46f2ffd99e19a89476e2f62270e0a... in tokenURI)
With that said, I went through some random NFTs the other day and I saw a lot of tokenURIs pointing to a plain metadata.json on a regular server that I could access in my browser.
The whole thing relies on a weird level of trust and belief in what other people believe to be valuable.
The problem as I see it is that NFTs lack any potential energy. Their value is purely kinetic, and expressed in the transaction. If no one wants my NFT, it's so much binary noise.
An NFT is just a pointer. Unless you have a contract specifically transferring additional rights, you literally just bought the pointer.
As money accumulates in larger and larger sums, and is abstracted farther and farther from tangible real-world power and resources, you get this. Things get increasingly weird.
There's got to be some kind of equilibrium point between this stuff, and your basic anti-fiat-currency gold bug, but I'm not sure what it is, particularly since modern macroeconomics acknowledges the capacity to just print money and make it up pretty quickly in the increased economic activity that makes possible: austerity isn't really a functional solution. This crypto stuff really works by the same principle: if you get enough people into it, they CAN'T crash because the important people are too deeply enmeshed in the system and will cheat to any degree to preserve the value of their properties.
I think it depends what is done with it. With macroeconomics, you can have entire countries spurred to activity and producing goods and services because they can transact with resources. With this crypto stuff, I very much wonder if it ends up being a small number of very privileged people demanding rights to increasingly silly abstractions that are said to be the value of entire cities, or countries.
I'm not sure that's sustainable, politically. I'd ask, how convenient is it for the first crypto trillionaires to buy real-world mercenaries? Things could get very dark, albeit in a peculiarly cyberpunk sort of way that might appeal to some.
So people might be trading (as in 2008) to "own" some N-th power representation of private debt that ends up going to zero because the underlying private debt itself was not sustainable. It's different from NFT, but for each degree of distance the financial instrument moves away from the real world, it looks increasingly weird, e.g. I own a share of insurance on a fraction of a bucket of debt people took out to buy their homes. (And I still think this is better than NFT unless the NFT has some underlying real-world thing tied to it.)
Maybe marginally better than URL but not very much IMO.
Vastly better than a URL. Content-addresses points to content that can be served from anywhere. URLs point to actual locations. Anyone can seed the content behind the IPFS hash, including the owner of the NFT.
While with a URL, you cannot change it. If the owner wants to make it work after it disappeared, they would have to buy the domain name, make the hosting work again and then setup a server there, add the file so it can be served.
While with a IPFS hash, the owner could just turn on their IPFS node and everything works as before.
Big difference.
If you're using a content-address hash, you can be sure that you can always get back the same result from that hash as when you got the NFT N years ago.
Is there a stat somewhere showing that IPFS is the most common kind of URL in NFTs?