The broken promises of NFT Art
NFT’s are often sold with the premise of permanence, but is the reality that most are destined to the trash heap of history?
After looking at the numerous options an artist has available to mint their artwork I found myself with more questions than answers. Originally I just wanted a platform that allowed the freedom to mint different filetypes (especially but not limited to .svg) and where I could get a little commission on the secondary market — a game-changing feature for artists embracing digital NFT’s — but as I eagerly delved face-first deeper into the options I found myself pondering for hours over the idea of persistence (I blame @j1mmyeth and his Zima red interview for this).
How can an artist guarantee an artwork will be accessible 5, 10, or 100 years into the future? What are the risks of HODL’ing an artwork for a lifetime? Can we re-assure art collector’s they’ll be able to pass NFT’s on to their siblings and they’ll inherit an artwork and not a worthless token pointing to a broken URL? I would love to mint the artwork directly on-chain but when a .gif can be >5MB this currently is too costly an option. On-chain artwork is still in its infancy, pioneered by larvalabs and the autoglyphs project, the image is stored within the token itself, making it as immutable as the blockchain itself. While the rest of the NFT world relies on off-chain files hosted externally.
Clearly, in an ideal world, all artwork should be minted on-chain.
Artists minting with platforms like Superrare, KnownOrigin & Rarible are placing trust in centralized platforms to stand the test of time and ‘do the right thing’ in terms of preserving content into the future. Some of these platforms are publishing to IPFS but it is unclear who is responsible for pinning these artworks (ie. ensuring they remain accessible) if they ever shut up shop. It’s my unoptimistic opinion that a good number of artists and collectors do not share these concerns about persistence and wish to make the quick buck before any of this becomes an issue. But who am I to judge, speculators increase liquidity, and maybe they’ll buy my artwork ;)
Here is a comparison of some of the popular minting platforms (details change regularly but are correct November 2020):
After looking at each of the options I found that none of these can be truly relied upon for persistent storage. IPFS is a perfect partner for NFT’s as each file is assigned a cryptographic unique content identifier (a CID hash) which can be used to prove the metadata and artwork’s authenticity (more info on CIDs here). But none of these platforms are able to guarantee that their servers and IPFS node will be up and running when we are six feet under pushing up daisies.
This curated platform is invite-only, you need to upload artworks, social media links, and a video to be considered. They used to store image data on their own servers but have now moved over to IPFS. Here is an example of a URL they use to retrieve a token’s metadata
If the Pixura IPFS node (Pixura seems to be a sister company to Superrare) ever went down, all links to NFT metadata and art would also.
Again this is a curated platform with similar requirements to Superrare. It uses the infura.io IPFS node to store and retrieve the external files required to view the artwork. Infura is owned by Joseph Lubin’s production studio ConsenSys, a big player in the ethereum space, but still not infallible. Just in the last week, Infura had a ‘major issue’ with its ethereum node provider service creating a huge headache for the entire ecosystem.
Rarible is the new kid on the block and doesn’t require approval to mint content. For that reason, there have been a few issues with bad actors stealing and reselling other artist's work. Rarible uses the daonomic.com IPFS node to retrieve artwork and this, again, is the weak link for their NFTs persistence.
There are some things you as an artist or collector can do to mitigate the risks of ending up with a broken NFT which I take a look at in my next article.
Just as my head was close to exploding I found the article “How to Create NFTs with IPFS and Digital Art” by Connie Digital. This quick guide boiled the complexity down to a simple workflow. Publish and pin artwork on IPFS using the Pinata service, mint your NFT’s with an OpenSea contract and point the external URL to the IPFS gateway. I can point to any type of file, and use a gif preview for the thumbnail to display within Opensea (this could result in issues when needing to load the higher quality files within platforms like Decentraland and Cryptovoxels).
Pinata allows up to 1GB of storage for free before charging users to pin. That cost I think should be down to the artist to manage, maybe the artist could reassure investors by putting aside a percentage of future sales to pay for IPFS persistence.
With this method the artist has control over uploading and pinning their art using Pinata, a service that lets IPFS know to keep it cached within the network, and has control over the URL used to access the IPFS network.
By using the ipfs.io gateway maintained by Protocol Labs, the well-funded team that developed the IPFS network (and other open-source protocols like Filecoin and libp2p), we can access content without relying on the Pinata node being operational in the future. Pinata seems like a reliable company with great prospects but compared to the organization that maintains the network itself I would argue the ipfs.io gateway is a safer access point to the IPFS network.
So you may choose to change your link from
(You still use the Pinata service to ensure persistence but this leaves options to change to a different IPFS pinning service in the future)
In Connie’s article, it briefly mentions the IPFS upload option “Preserve filename (wrap with directory)” and that it’s required in order to maintain the same filename across IPFS. But is it actually important to preserve the filename within IPFS? In my opinion, the IPFS hash proving the metadata & artwork’s legitimacy is more important than preserving a filename. After all, the artwork title is included in the metadata, and in fact, artists naming their own works at all is a fairly recent phenomenon. For hundreds of years, art historians used descriptions instead of official titles to identify specific works, and now IPFS allows us to cryptographically identify a digital image with it’s CID.
It's still possible to find the image CID from the first example however it's actually included in the URL in the second example (highlighted in blue), which in my view makes it more meaningful and less confusing for people unfamiliar with IPFS.
You can test this yourself by downloading the image at https://ipfs.io/ipfs/QmP43Y129T6KYFu9neaDLvfrF5HJmP9UueBZa3LQ1sugvS and using the Pinata CID checker.
The cool thing about this approach is if your artwork ever gets de-pinned in the future, literally anyone with a copy of it can make the NFT whole again by repinning it to the network (*cough* — as long as the ipfs.io gateway were still around…). So I encourage artists and collectors to think about regularly backing up artworks as well as their metadata files or at least to pin the metadata and artwork themselves to ensure robustness on the network.
This approach to minting seems great! What's the catch?…
Opensea does not store metadata on IPFS but instead choose to on their own Opensea servers. I can understand why, NFT’s used for games may need to frequently update a token’s metadata, this just wouldn’t work on IPFS as the CID reference hash would change on every update. However, it is quite annoying for artists concerned about longevity over flexibility. It looks like for now, without writing your own Solidity contracts, artists have a limited number of options for minting, all of which could break in the future.
It's early days but the future looks golden.
This space is evolving at break-neck speed, and the concerns expressed by Jim seem to have been heard by the community. Just in the last few days, InfiNFT.com has opened to the public, a new NFT minting platform writing all metadata on-chain and using both IPFS and Arweave for content storage. This is a great step in the right direction for NFT land and I’m looking forward to testing out the platform, but with filecoin and other decentralized storage solutions just around the corner, it’s clear the holy grail of permanent digital NFT art may be only a matter of months away.