Computing desk
< March 22	<< Feb \| March \| Apr >>	March 24 >

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

March 23

NFT data files

Non-fungible tokens seem to be associated with images (or sometimes other kinds of data file) only in a certain way: the internet address of the data, I think, is put on the blockchain. The data itself is of no importance (and is typically underwhelming, or ridiculous) but if I understand correctly the internet address acts as a unique identifier, just like the naming conventions for Java packages, where an internet address is used to make sure nobody's code (which they might want to share) accidentally has the same name as anybody else's. In the case of Java, the package doesn't have to be actually located online at the given address, or anywhere, and the complete address doesn't even have to exist, but the "organisation" (or lone bedroom coder) creating the code is supposed to own the internet domain in order to ensure that nobody else's code gets the same name. So, what I'm asking is, why can't NFTs work the same way as Java package names? If the only important thing is that the token is scarce, why can't the token simply be a web address (which might not even need to be more than potentially real)? Or a UUID? What's with the great urge to seek crappy pixel-art images to "turn into NFTs", with people even resorting to plundering them from other people's DeviantArt pages (according to our article), when the connection to the actual token is so tenuous? Card Zero (talk) 02:10, 23 March 2022 (UTC)[reply]

I don't think you even need an internet address. Just an entry in the blockchain that says "this object belongs to the user with the secret key to the public key XXX". It's the purest example of artificial scarcity that exists. It scarcity of something completely useless and not even scarce. And we burn dead dinosaurs to maintain this non-scarce scarcity. If that is not a sign of the end times, I don't know what might be... --Stephan Schulz (talk) 22:01, 23 March 2022 (UTC)[reply]

The entry needs to contain some digital description of the object, like "object YYY belongs to the user with the secret key to the public key XXX". Nothing by itself will keep me from adding another entry "object YYY belongs to the user with the secret key to the public key ZZZ", so an unforgeable time stamp or equivalent is required to establish precedent. Even then, what keeps me from entering such an entry preemptively, even though I have no say whatsoever over the ownership of object YYY? --Lambiam 22:34, 23 March 2022 (UTC)[reply]

I don't know. I had only got as far as vaguely realising in the back of my mind that this sort of fraud would be possible, and the timestamp seems like a solution, and then you immediately find a hole in the idea. Maybe the issuer of the token also has a key which can be used to confirm that the token is genuine? Card Zero (talk) 00:07, 24 March 2022 (UTC)[reply]

I think the planet's Ether miners collectively fight your attempt to change the art token owner till you give up, and whoever has more exaflops wins cause they can extend the gigabytes of chained blocks faster than you in the long run even if your hash guess(es) are really lucky for the first few blocks, and the longest chain is considered the real one, and every so often one of the miners gets free crypto, which they can sell for real money, I think the blocks are newspaper snippets or brags or ant names or porn URLs or recipes or spam or anything else the successful miner wants to immortalize/force the universe's Ether users to download for the rest of time, and/or the record of which virtual wallet numbers have more or less Ether than at the last block due to payments, and maybe other stuff, all in binary and in a standardized format set at the format's creation, and each miner guesses the binary string that'll hash to the block or the part that's hashed, as many times as their single-purpose mining supercomputer or GPU etc can guess per second and the luckiest guess gets a predetermined amount of free Ether, which wasted an amount of electricity worth a large percent of the Ether's price in real money to find, cause the anarchist system somehow adjusts to keep the free cryto poop rate near one per x minutes no matter how many exaguesses per second and gigawatts are thrown at the virtual mine Sagittarian Milky Way (talk) 03:11, 25 March 2022 (UTC)[reply]

It could be (and I'm only guessing) that fashion and prestige associated with certain images (bizarre though it seems for a scrap of pixel art to be prestigious) means they can command a higher value and seem more scarce, whereas an unmemorable address alone (or ID number, or key) would seem just like all the others and hence not scarce at all. Even so, I get the impression (again just a guess) that most of the market consists of thousands and thousands of highly unremarkable images: but again, perhaps the point is to at least minimally distinguish the item from a mere number. The situation could be just like the physical art market: a bland watercolor of a beach scene might be one of millions of effectively identical items, yet still be worth more than a mere six-digit number (since, perceptually, those are all fungible), while at the same time being worth less than a stylish abstract work painted in bird droppings by a deceased artist with a tragic backstory, which would really be perceptually near-unique, even though technically the mere number is equally as unique. Card Zero (talk) 00:02, 24 March 2022 (UTC)[reply]

An NFT as commonly understood is an Etherium Smart Contract implementing ERC-721. This standard doesn't require any kind of description, but if it does contain (optional) metadata, it should contain a URI according to that standard. And it's suggested that URI points to a JSON object which points to an image. El sjaako (talk) 15:58, 24 March 2022 (UTC)[reply]