Provenance. The Romainiti case
Series introduction
This research examines, from the perspectives of curatorial practice and collecting, what remains of tokenised digital art following the collapse of the NFT speculative market. The series is developed in three entries that articulate three dimensions of a single problem: provenance, persistence and recognition. This, the first in the series, focuses on the former two through the archive and the analysis of a specific case.
El antecedente inmediato de esta investigación es la entrada “Panorama del mercado global del arte”, publicada en este mismo Cuaderno público, que sitúa el contexto macro: contracción brusca del mercado de Art NFTs (de 2.900 millones de dólares en 2021 a 23,8 millones en el primer trimestre de 2025, según DappRadar), absorción institucional creciente del arte digital, y separación cada vez más nítida entre el arte digital como medio y el NFT como infraestructura.
The framework that articulates this research—provenance, persistence, recognition—is not a speculative proposal. It is the conceptual formalisation of criteria that the author has applied since 2021 in his professional practice, particularly in the expertisation protocols for native digital art developed at Kripties. The public background to this position is documented in the corresponding section of this website.
This research is authored by Juan A. Esteban. The case developed in this entry corresponds to the Esteban-Ruiz Collection, the author's personal collection as a private collector.
The problem of the archive in digital art
In material art, the archive is a late-stage concern. The work exists first as an object—a canvas, a sculpture, an installation—and only later, when it must be conserved, catalogued, loaned or inherited, does the archival question arise. The museum, the inventory and the certificate of authenticity arrive at a piece that already possesses a body. In native digital art, the sequence is inverted. The work does not pre-exist its archive: it exists only insofar as a storage system, a blockchain, a contract and a series of metadata sustain it as an identifiable entity. If that infrastructure fails, it is not the conservation of the work that fails; it is the work itself.
This inversion alters the nature of the problem. In material art, one may debate whether a restoration is excessive, whether an attribution is correct or whether a provenance is documented, but the physical piece usually remains present while the debate continues. In digital art, a server outage, an unrenewed domain or an IPFS node without pinning are sufficient to ensure that what is called into question is not the attribution, but the very availability of the piece. The work does not physically degrade: it becomes irretrievable instantly, silently, without a material trace.
Consequently, provenance in this field cannot be understood as an isolated technical datum—the token hash, the minting transaction, the author's wallet—nor as the digital equivalent of a gallerist's certificate. Provenance is the sum of several layers: the on-chain record, the effectively accessible archive, the metadata linking token and file, the curatorial context that stabilised the piece as a work, and the sustained capacity of the author or collector to maintain all of this over time. Each layer possesses its own regime of fragility. The chain is practically permanent; the archive depends on the storage model; the metadata depends on URLs that may fail; the curatorial context depends on pages, platforms and narratives that may disappear; human sustainability depends on economic, technical and geopolitical resources that are rarely symmetrical between the parties involved.
Treating provenance as a dimension, rather than as data, necessitates evaluating each layer separately and simultaneously. A work may possess an impeccable on-chain record and a lost archive. It may have the archive and metadata in order, yet have been decontextualised by the disappearance of the platform that presented it. It may have all of the above and still be left orphaned if the artist loses the technical or material capacity to maintain it. Provenance is sustained when all layers are sustained; it degrades the moment one fails, even if the others remain intact.
Thus, the archive in digital art is not the final phase of the process—the conservation of something previously given—but the primary condition of the work's existence as an identifiable piece. Before asking whether a piece is valuable, whether it is correctly attributed or whether it has a market, one must ask whether it is there, whether it will continue to be there and under what conditions. That question is not answered by looking at the token in isolation. It is answered by looking at the complete ecosystem that sustains the token: chain, archive, metadata, context and maintenance capacity. When a platform falls—and they fall, frequently, without noise—the question ceases to be theoretical: what survives and what does not?
Technical anatomy of the NFT
An NFT is not, strictly speaking, an image, a video or a digital work. It is a record on a blockchain that represents the ownership of, or the relationship with, a digital asset. The distinction is not rhetorical: the NFT is the token, not the file. Understanding that separation is the condition for understanding why some NFTs survive the fall of a platform and others remain empty.
The token resides on the blockchain (Ethereum, Base—which operates as a layer 2 on Ethereum—or other networks) as a unique identifier managed by a smart contract that stores the token ID, the owner's address, and a reference to the metadata in the form of a URI. The file almost never resides on-chain: it resides externally, and the token points towards it. This asymmetry introduces a structural division between what is on-chain—ownership, history, contract logic—and what is off-chain—file, metadata, hosting. The actual persistence of the work depends, above all, on how that second layer has been resolved.
There are four standard methods for hosting the work to which a token points, and it is advisable to present them in order of decreasing technical sovereignty. The first, and the only one that does not introduce external dependencies, is fully on-chain: the content—image, generative code, text—resides within the smart contract bytecode itself, encoded on the chain. In this model, there is no external file to point to: the token and the work are one and the same, and as long as the chain exists, the work exists. Projects such as Autoglyphs, Avastars, or a significant portion of Art Blocks Curated operate in this manner, generally under considerable technical constraints—the cost of gas necessitates lightweight pieces, almost always generative code or minimalist graphics—but offering in return the only form of persistence that requires no external actors.
The other three modalities introduce, in different ways, this external dependency. The second is traditional platform servers, typical of the Web2 model: Foundation, OpenSea or, in part, KnownOrigin. The token points to a conventional URL; if the platform fails, the file ceases to resolve. The third is IPFS (InterPlanetary File System), a distributed system in which files are identified by their hash, not by a fixed URL. Content ceases to depend on a single server, but it introduces a trap: pinning. If no one keeps the file active on the network, it may disappear. IPFS distributes; it does not archive. The fourth is Arweave, which offers perpetual storage under a single payment model with an economic incentive to maintain data in the long term. It is the option closest to real persistence outside of the fully on-chain model, although its initial cost is higher and not all projects adopt it.
In addition to the question of the file, there is the matter of control. An NFT may have been minted on a contract shared by an entire platform or on a contract owned by the artist, and the difference has structural consequences. In the shared model, common on Foundation or KnownOrigin, many artists coexist within the same smart contract, and the token logic depends on the company's infrastructure. In the own-contract model, frequent on Manifold and partially on Zora, the artist deploys their own contract and retains total control over tokens and metadata. Portability is improved, and dependence on the platform's fate is reduced. Persistence, therefore, is not merely a technical question of storage: it is also a question of governance.
Standards introduce a second variable. ERC-721 associates a token with a single unique work; it is the classic model, clear conceptually. ERC-1155 allows multiple copies under a single ID, combines fungible and non-fungible logic, optimises gas costs, and proves more efficient for series and editions, although it modifies the logic of scarcity.
When a platform fails, it is advisable to separate the layers. What is on-chain is not lost: the ownership of the token, the transaction history, the smart contract, and the relationship with the wallet remain as long as the chain exists. What may be lost lies outside: the file if it depends on servers or an IPFS without pinning, the metadata if it points to broken URLs, and, above all, the curatorial context—the page, the collection, the narrative that sustained the work as an exhibited piece. In the long term, only accessible contracts, recoverable metadata, and effectively persistent files survive. The rest degrades.
There are community tools for interacting with legacy contracts from platforms that have reduced their activity, such as savemyart.xyz, used by collectors and artists to manage old Foundation listings. It is advisable to cite these with caution: they do not have a publicly identified operator nor do they constitute official infrastructure, so their use requires technical verification and reasonable precautions—secondary wallet, review of transactions on Etherscan, and not blindly signing a setApprovalForAll whose implications have not been understood. The KnownOrigin case illustrates the broader scenario: acquired by eBay in June 2022, it subsequently reduced its activity and public visibility. The works minted continue to exist on Ethereum, but were partially subordinated to the fate of the platform that hosted them.
The true persistence of an NFT work rests, therefore, upon three layers: the blockchain, where the token is practically permanent; the storage infrastructure, which determines whether the file lives or evaporates—except in the fully on-chain model, where this layer coincides with the first; and the control model, which determines the actual capacity of the artist or the collector to sustain the work over time.
The Case. Relocation from Foundation to Transient
Foundation was, for years, one of the most respected invite-only NFT platforms in the Ethereum ecosystem. Selective, curated, with a polished visual identity. It sustained the work of hundreds of artists during the expansion cycle. In April 2026, the platform announced its closure after failing to find a buyer, although it will continue to support IPFS data for one year.
What follows is an account of a specific case. The artist romainiti, based in Iran, with work on Foundation and a profile now on Transient (https://www.transient.xyz/@romainiti), agreed to share her experience for this investigation. Eleven of the works minted by romainiti on Foundation remained in the possession of Juan A. Esteban as a private collector, integrated into the Esteban-Ruiz Collection, which also includes work by the artist from other platforms. The voice that follows is hers, translated from English with her authorisation, maintaining the literal meaning of her contribution.
How she found out
Se enteró por publicaciones de otros artistas quejándose en redes. Su reacción inicial, en sus palabras: “honestly I didn’t give a little fuck because I had more serious problems going on in my country and my life couldn’t think even about a market place”. Hacía tiempo que no tenía ventas allí.
However, upon confirming that Foundation was closing completely, she did begin to worry. Not for herself in the abstract, but for the works and for her collectors. She began to investigate what to do.
The decision to move and why Transient
She had works on different networks—Ethereum and Base, primarily—some sold, others in her possession. The choice of Transient was based on three specific criteria: the platform supports both Ethereum and Base; it offered a specific plan for Foundation artists that allowed for locating collections, importing them, and re-listing unsold items; and it maintained an artistic focus, rather than that of a mass marketplace.
The technical process
It involved three steps: connecting the wallet to a third-party site and locating the works, withdrawing them from the original listings, and, subsequently, importing the collection to Transient and re-listing. The intermediary tool used was savemyart.xyz, a community tool from the post-Foundation ecosystem that allows for scanning the wallet, locating works still listed on old contracts, and cancelling those listings.
En palabras de romainiti: “to be honest it wasn’t clear at all and it was so confusing for me and I was so mad at foundation for causing me too much trouble”. Una colega artista y coleccionista, Jennifer, la acompañó en el proceso.
The real cost, in her context
Here the case diverges from any purely technical description. Iran has been under internet blockade and recurring blackouts since the beginning of the Iran-United States conflict. The connection, in practice, depends on purchasing VPN packages. Before the blockade, 100 GB cost approximately 1.5 dollars. Afterwards, each GB costs between 5 and 10 dollars. The exchange rate of the Iranian rial against the dollar continues to fall (at the time of the process, approximately 1 IRR ≈ 0.0000007601 USD).
Importing the works and re-localising them consumed 3 GB of VPN. All the VPN purchased the previous day was completely exhausted and had to be purchased again. To this was added the cost in ETH of the gas fees for re-listing. They used all the ETH they had.
What was left behind
Some editions held on Foundation could not migrate to Transient as editions. They were forcibly left as unique pieces (1/1). For those works, sold previously at the price corresponding to an edition, the change in status had no retroactive effect: they were already in the possession of whoever had acquired them, now with a status different from the original. In the case of works acquired by the Esteban-Ruiz Collection, Juan A. Esteban was the sole buyer of those editions, which allowed the conversion to 1/1 to occur transparently, without affecting other collectors.
What the collector saw
From the perspective of the private collector, the experience was radically different. The artist communicated the situation, managed the entire process, and required no technical intervention. It sufficed to log in to Transient with the same wallet associated with the user account on Foundation. The works appeared automatically assigned, now visible from the new interface. On-chain ownership, guaranteed by blockchain, remained intact. For a collection maintained with a curatorial rather than commercial vocation—the Esteban-Ruiz Collection is built without the intention of commercialising the works—the change was, in practical terms, invisible. The work remained the work, with the same traceability of ownership.
The asymmetry, however, is significant. What was invisible to the collector was, for the artist, a costly, technically confusing operation dependent on peer support. That asymmetry is not accidental: it reflects how the real cost of keeping a digital work alive is distributed when a platform fails.
Foundation closure
The final advice from romainiti, in direct English:
“Make sure you know your art contract, you have it on blockchain. This happened to KnownOrigin, Foundation… and can happen again, so make sure your art is safe on blockchain.”
KnownOrigin—an NFT platform acquired by eBay in June 2022 and subsequently reduced in its activity—is a precedent that the artist names explicitly. It is not the first time they have experienced a closure. It will probably not be the last.
What the case teaches about the framework
The romainiti case allows us to return to the triadic framework through experience, rather than theory. Provenance, viewed in light of what occurred between Foundation and Transient, ceases to appear as an attribute of the token and reveals itself as what it effectively is: a fragile construction sustained by heterogeneous layers that rarely fail in unison. When one of these layers breaks—the file, the metadata, the curatorial context, the maintenance capacity—provenance degrades even if the token continues to exist on the chain. The work may remain registered and, nevertheless, have lost the ecosystem that rendered it legible as a work.
The movement of the eleven pieces was not an on-chain migration: the original contracts remain where they were, and what changed was the platform that indexes them and allows for their operation. Technical precision matters, because it renders the case interesting. We are not witnessing a transfer of tokens, but a recontextualisation: material provenance is maintained; operational provenance—that which allows a work to circulate, be viewed, and be recognised as such—is reconstructed in another environment.
Hence the dual decision that articulated the process. Transient functioned as a retroactive rescue mechanism for works already minted on Ethereum and Base, thanks to its specific plan for artists coming from Foundation. However, future continuity was decided in another ecosystem: OBJKT, on Tezos. The choice was not arbitrary. Tezos has consolidated its position as one of the standards with the greatest implementation in the part of the digital art world that takes economic and energy sustainability seriously; the minting cost is low, which allows artists with limited resources to operate without every technical decision becoming an economic dilemma, and the stability of the ecosystem reduces exposure to the closure or reorientation of platforms. Rescue on one network, continuity on another: two distinct responses to distinct problems, not a contradiction.
El caso también desplaza el centro de gravedad de la discusión. Si la procedencia depende de capas heterogéneas y la capacidad de sostenerlas es desigual entre artista y coleccionista, entonces la pregunta importante no es solo técnica. Es una pregunta sobre criterio. Y aquí conviene citar literalmente al coleccionista que vivió el proceso desde dentro: “no me preocupa, sino que me reafirma en lo importante que es la curaduría: adquirir obra de artistas fiables y contrastados”. La frase condensa el desplazamiento que esta serie quiere proponer. La fragilidad de la infraestructura no se compensa con más infraestructura, sino con un trabajo previo de criterio: qué artistas, qué prácticas, qué trayectorias soportan el tipo de relación a largo plazo que el arte digital exige a quien lo colecciona. La procedencia, en última instancia, no la sostiene sola la cadena: la sostiene la decisión informada de seguir a determinados autores a través de los entornos en los que se mueven.
The dimensions of persistence and institutional recognition, which the triadic framework has barely enunciated until now, open questions that the romainiti case leaves on the horizon unresolved: how the durability of a digital work is constructed beyond a one-off rescue, and to what extent institutions—museums, public collections, expert bodies—are in a position to recognise and sustain a heritage that operates under logics distinct from those that have historically organised their work. These are the questions opened by entry II.
Sources
Platforms and tools mentioned
Foundation, https://foundation.app
Transient, https://www.transient.xyz
savemyart.xyz, https://savemyart.xyz
KnownOrigin, https://knownorigin.io
OBJKT, https://objkt.com
Manifold, https://manifold.xyz
Zora, https://zora.co
OpenSea, https://opensea.io
Etherscan, https://etherscan.io
Technical infrastructure
IPFS — InterPlanetary File System, https://ipfs.tech
Arweave, https://arweave.org
Ethereum, https://ethereum.org
Base, https://base.org
Tezos, https://tezos.com
ERC-721 — non-fungible token standard, EIP-721, https://eips.ethereum.org/EIPS/eip-721
ERC-1155 — multi-token standard, EIP-1155, https://eips.ethereum.org/EIPS/eip-1155
Market data
DappRadar, Industry Report Q1 2025. Art NFT trading volume: $23.8 million.
El País, “De valer millones de dólares a no llegar a las cuatro cifras: la gran ruina de los NFT”, 10 de marzo de 2026, https://elpais.com/icon-design/2026-03-10/de-valer-millones-de-dolares-a-no-llegar-a-las-cuatro-cifras-la-gran-ruina-de-los-nfts.html
BeInCrypto, “Desplome del NFT: cambio, colapso, cofundador de Rarible”, https://es.beincrypto.com/desplome-nft-cambio-colapso-cofundador-rarible/
Background
“Panorama del mercado global del arte”, Cuaderno público, https://juanesteban.art/panorama-del-mercado-global-del-arte/
Background to the Kripties Foundation position, corresponding section of this website, https://juanesteban.art/investigacion/antecedentes/
Case
romainiti profile on Transient, https://www.transient.xyz/@romainiti
Personal communication from the artist to the author, [May 2026]. Quotes reproduced with authorisation.
Acquisition of KnownOrigin by eBay (June 2022), https://www.ebayinc.com/stories/news/ebay-acquires-leading-nft-marketplace-knownorigin