Rick Manelius, PhD, is a nanotech engineer turned serial startup co-founder. His current role is CEO of Atomic Form, which builds multi-chain provenance solutions for NFT creators, collectors, and curators. He is passionate about a variety of tech topics relevant to Web3, including metadata standards, syndication, publishing, playback, and long-term archival of digital assets. He is a Techstars all-star mentor (an honor held by less than 2% of the thousand of mentors in the worldwide network). His book, Save Your Startup, is a series of playbooks for first-time founders looking improve their odds of success. Rick can be found at http://rickmanelius.com or on Twitter at @rickmanelius
No comments yet. Be the first to share your thoughts.
The longevity of digital art is under threat. Without a major push to standardize NFT technology, we will eventually face a future in which today’s vibrant new world of digital art, with all its value, could suffer catastrophic loss (we’ve already lost priceless works by XCOPY). Only a unified approach to NFT technology across the industry can ensure that over the long term, artworks and entire portfolios do not dissipate into the digital ether. It is crucial to draw timely lessons from the dead ends of Web1 and the history of analog art preservation. Below, we examine the stakes and issues in detail, as we advocate for a solution to anchor the value and accessibility of NFTs for generations.
“Your NFT displays are broken.”
NFT collectors are faced with NFTs that won’t render on many displays, and often attribute this problem to the display manufacturer.
The actual cause lies deeper. NFTs are not just JPEGs. There is a lot of potential complexity behind them. Content size demands storing large media files outside the scarce and expensive blockchain storage space, linked together by fragile location descriptors (a.k.a. URLs).
If off-chain NFT media files get relocated, links can, and do, break. The NFT is robbed of the content it represents.
Web2’s shortcomings led to media fragmentation. It’s a significant issue for NFT creators and owners.
But at the end of the day, collectors don't care. They just want their displays to work.
It’s not easy to be the bearer of bad news.
To create future-proof, long-living NFTs, we have to agree upon standards for the two fundamental requirements for NFTs:
Playback, or the ability to render the media of an NFT as the creator intended
Syndication, or the ability to transfer, license, sell, or publish an NFT
Playback requires standardized data, while syndication requires standardized metadata. Both are essential for the structural integrity of NFTs.
NFTs and digital art have a high transformative potential, and we must preserve this potential for the foreseeable future. Web3 needs unified standards for both content and metadata to make digital art accessible for decades and centuries to come.
History repeats itself. Like the NFT space today, the era of Web1 also struggled with fragmentation and lack of standards. Web content would often only render correctly on Internet Explorer, the dominant browser of that time. Playing out sophisticated, interactive content required using a proprietary (and bug-ridden) browser plugin called Flash Player.
Internet Explorer and Flash are long gone. Content that was built for these specific products lost its ability for playback (even if you can access the original files from the Internet Archive).
Web2 learned from the mistake of not standardizing content. Today, most web content you encounter adheres to some well-known standard format: JPG, PNG, WEBP, MP3, AAC, MP4, M4A, HTML5, CSS3, PDF, SVG, and so on. Software strives to stay backward-compatible with old standards.
Moreover, Web2 stands for another substantial achievement: the standardization of metadata. While content is data, metadata is information about this data. Digital photographs carry EXIF metadata, documenting when and where a shot was taken, what camera model and lens were used, the image's resolution, and so on. Web pages include invisible “<meta>” tags that reveal the author, creation date, description, and more. With standardized metadata attached, content became easy to syndicate. Web2 established the power of “Create Once, Publish Everywhere” (or COPE), which allows repurposing the same content across multiple channels.
Remember the issue of NFTs playing on OpenSea, but not on some NFT displays? This is a problem of syndication, ensuring that the playback can occur across a variety of apps and devices.
Yes, history repeats itself, and by taking our learnings from Web3’s repetition of the unsuccessful parts, we can initiate the repetition of the successful part.
Back to the present state of Web3. The NFT era is about to reach dead ends like Web1 did. Like Flash, some smart contracts are proprietary and wouldn’t work outside the company’s context. Like the uncontrolled growth of video codecs, novel contract types abound, yet only a few of them reach mass adoption. Like early HTML, NFTs have no standardized metadata formats. Where is the equivalent to the COPE pattern for NFTs?
Looking at the current NFT landscape can make your heart sink. It is the same fragmented landscape as in Web1: multiple chain technologies, numerous NFT platforms, proprietary marketplaces, and inconsistent viewing experiences. Fragmentation is an outright threat to interoperability and composability. Analysis of NFTs frequently reveals common issues like broken links or incompatible metadata schemas.
Web1 should have already taught us that inadequate metadata can cause severe issues with compatibility, presentation, and content sharing. Early search engine result pages relied on Google scraping a page and guessing its summary. Even when HTML metatags were introduced, people didn’t know how to use them. Too often, metadata is treated as a blank canvas that offers flexibility and an open field for experimentation but without any guardrails.
Today’s NFT situation is even worse than in Web1. A website is a living object that can be changed any time to adopt new features and standards. NFTs are the exact opposite. In many cases, once an NFT is minted, all of the metadata included in the minting process is frozen forever. There is no way of correcting on-chain NFT metadata or amending it with new data.
Earlier, I mentioned COPE, a concept that ultimately frees content from being restricted to particular distribution channels or presentation formats. The NFT world must embrace a similar approach to overcome fragmentation. We propose a concept called MODE: Mint Once, Display Everywhere.
The minting process is the starting point of any NFT. The data and metadata included in the minting process decide the future of an NFT.
Can metadata be changed or appended?
What legal uses are specified for licensing and remixing?
Does the NFT include multiple formats for specific devices or endpoints?
What about future developments? Can the NFT provide formats ahead of time? What defaults and settings might a future application expect? Think of recent developments like VR headsets or LLMs that display or consume content in new ways.
Does a generative or dynamic NFT provide sane fallbacks (like a static image or a video) if a device does not support the rendering algorithm?
Making NFTs future-proof requires a new approach to defining NFT content and metadata.
Unlike classic paintings that exist on a single, fixed-size “output device” (a canvas mounted on a wooden frame), digital art comes to life on different devices. Therefore, NFT content must be encoded in a format suitable for display on any device. There is no need to invent a new format, however. Today’s media technology already provides flexible content formats. Still, we need to be careful to choose the ones that are already a widely accepted standard and likely to be supported in the future.
A unified content format also needs to include possible fallbacks for devices incapable of playing back the original format. This is not just a theoretical concern. Take Chromie Squiggles, for example. Chromie Squiggles are generative art that renders in real-time on an HTML canvas. Underpowered hardware (like a Raspberry Pi 4) may have trouble displaying a Squiggle at the original speed. To overcome this problem, NFT formats that use live rendering could include a pre-recorded video loop as a fallback.
Web2 has achieved a lot in this direction. Think of HTML5 and the introduction of semantic tags (<article> instead of <div class=”framework_defined_tag_for_article”>) that make HTML5 content universally playable across different media, from browsers to e-books or screen readers. Think of video streaming services that adapt video resolution to the available bandwidth. Think of the many W3C standards that just about every modern browser implements, eliminating the need for proprietary plugins.
In short, unified content ensures interoperability for playback.
While content formats define playback options, metadata defines syndication options. Web2’s COPE would not have been possible in any meaningful way without metadata standards (which include de facto standards) like RSS, Atom, Schema.org, or Open Graph.
NFTs are routinely syndicated to marketplaces, social media, artist platforms, dedicated NFT art platforms, crypto art galleries, art fairs and hardware displays at exhibitions, and many more places. Only if we standardize NFT metadata for universal consumption by syndicators can we leave the world of isolated NFT islands behind.
In short, standardized metadata ensures interoperability for syndication.
The world is full of contradictions. Artists want to create art that outlives them and that withstands the test of time. Yet, each piece of art collects ongoing lore of its activities and stories around it while moving from one owner to another, being exhibited at art galleries or discussed in blogs or on social media.
All this information is worth connecting to this piece of art. For NFT art, the best place for this information is the NFT metadata. Yet, if metadata is added during the minting process, it is frozen forever, and if it is merely associated with an NFT after minting, no one can prove the authenticity of this data.
So when we talk about standardizing NFT metadata, we have to consider its dynamic nature. In addition to standard formats, we need a standardized, chain-independent way of attesting that a given metadata entity belongs to a particular NFT.
Formally, we can distinguish between two forms of updating metadata:
Off-chain metadata is mutable and can be directly updated. To link the new metadata back to the original NFT, the metadata and its relationship to the NFT must be attested through cryptographic signing.
On-chain metadata is immutable by definition. Here, updating metadata means appending new metadata with appropriate attestation.
So by creating a cryptographically provable link between an NFT and updated metadata, we can achieve the (almost) impossible: to create or append new metadata for an existing, immutable NFT.
Content and metadata standardization solves many problems, but the resulting benefits are more than the sum of problem solutions.
A seamless user experience. NFT creators, owners, curators, and consumers will have a seamless experience across playback devices, chains, and other contexts.
Preservation of digital art for future generations. The provenance and accumulated history of a digital piece of art is as essential as the artwork itself. Standardization ensures that future generations can enjoy digital art as much as we do now.
Easier distribution and syndication. NFT art will be able to escape the boundaries imposed by incompatible blockchain technologies, metadata formats, and chain-specific attestation solutions.
Better presentation of NFTs. Digital art will render in the best quality on today’s devices and devices yet to be invented (Apple Vision Pro came out just last year).
None of this will happen if the NFT community does not take action. Even worse, taking no action means the situation will deteriorate further. We need to make an ongoing effort to reverse entropy.
We must create unified content and metadata structures where only individual, narrowly focused, isolated solutions currently exist.
And we have to start in multiple places to get the ball rolling.
Creators, owners, and curators need to become very aware of the importance of high-quality metadata standards. Without a conscious mind shift, humans tend to go on forever in the way they always did. But humans also know how to trick their own brains into escaping this trap.
Marketplaces and smart contract authors must choose to adopt sane defaults around metadata. Examples from the past repeatedly show that this is possible. When online image galleries like Flickr gained widespread popularity, copyright infringements were a common problem. As a result, we now have standardized licenses like the family of Creative Common licenses that content creators can select to protect their work while making it available to others under well-defined conditions. Another example is Facebook, which had to deal with syndicating external content of any kind into their users' timelines. Facebook addressed this challenge by establishing Open Graph, a de facto metadata standard for syndicating content to social media.
Lastly, we need more initiatives showcasing existing NFT content and metadata standardization approaches. One potential solution developed by my team called Atomic Lore shows that cross-chain attestation of NFT metadata is entirely possible today by combining immutable and verifiable off-chain storage systems (IPFS and Arweave) with chain-independent, wallet-based signatures.
NFT artwork is precious, often worth thousands of dollars, but the current state of affairs is alarming. Without future-proof standardization of content formats and metadata, the value of an NFT may decay over time. Today, an NFT may play flawlessly. But the future may hold nasty surprises if NFTs continue to rely on experimental live rendering techniques that may or may not be supported by tomorrow’s playback devices.
Fragmentation happens naturally and inevitably unless the affected parties become aware of this fragmentation, with all of its downsides, and decide to actively change course. It is upon us, the NFT creators, owners, curators, syndicators, traders, and consumers, to decide if we want to build more islands, more dead ends, or if we want to build a unified NFT experience for everyone.
The NFT community must develop a massive interest in arriving at agreed-upon, meaningful content and metadata standards, for the love of digital art.
Mentioned in this article
People: Rick Manelius, XCOPY|Works: Chromie Squiggles|Organizations: Facebook, Google, OpenSea
