The Ordinals Ecosystem is currently undergoing some soul searching for how creators, marketplaces, and the protocol itself should recognize inscriptions as falling under a collection or not.

Contrary to NFTs on Ethereum, Bitcoin does not have a native contract structure you can use to check an asset’s provenance. Early on, most inscription IDs were tracked by creators as they were inscribed. This JSON list of all the inscription IDs would typically be pinned in a Discord channel and manually passed around to the numerous marketplaces popping up. 

As time went on, Casey Rodarmor eventually added the parent-child demarcator to the Ord client, an enshrined provenance scheme (that he notably uses for his first collection, FUN!). Additionally, Casey and the team added metadata to the Ord client, although most inscriptions do not utilize this, nor has a metadata convention emerged for tracking collections.

Casey devised the Ord client’s method for expressing an inscription’s provenance as an alternative to Ethereum’s contract verification method (and one of the only workarounds that could actually work). 

“One of the things that Casey didn’t like about ETH NFTs is that for each NFT, there was a smart contract that everyone has to verify….Now Ordinals is great because it’s one protocol. So in this one protocol, there are limited ways to do things, and Parent-Child Provenance is one of those limited ways.” – Danny Yang

Soul searching for ordinal provenance

A year and a half after Casey’s parent-child scheme, though, the Ordinals ecosystem continues to struggle with how marketplaces, explorers, and creators should treat provenance. Today, we have a couple major points of friction for the ecosystem.

1. As the ecosystem has grown, a single marketplace has come to dominate 90%+ of Inscriptions trading. MagicEden has essentially become the single reference point for collection data for the entire Ordinals ecosystem. This has placed them in a gatekeeper role and collection success pretty much lives and dies by whether the team gets it listed on the platform.

Is it an existential risk or structural barrier to healthy ecosystem growth if a single marketplace wields this much power?

2. Let’s say someone loses inscription in a hack and those inscriptions are listed on a marketplace. In a common response, the collection founder removes those inscription IDs from the official roster. While not actually affecting what is on chain, this creates the second order effect of permanently removing these items from any lists at all and may result in their being memory-holed. 

The first prominent incident of this dynamic occurred with the Pizza Ninja collection, when the team removed the inscription ID of a ninja that was confirmed stolen. MagicEden removed that inscription ID from the official collection recognizance. This has been a common approach for many collections over the past year as they seek to protect holders who have been scammed or hacked.

Minority share marketplaces such as Magisat often find themselves pressured to match MagicEden’s decision to remove the inscription ID. Recently, Magisat founder Const_Quary made the decision not to remove several stolen items from the Based Angels collection from the marketplace’s index for the collection, prompting criticism. As a result, this collection’s record is fragmented across markets, with two markets indexing the collection in different orders. This fragmentation also allows one marketplace to exclusively order a collection while the others scramble to emulate this order (usually without access to any kind of automation or API).

Casey believes these consequences are a result of not releasing parent-child provenance earlier. In an X post, he laments: “I definitely regret not getting parent-child provenance out with the first release of inscriptions. I should have known that everyone would immediately start using JSON-based provenance.”

Is removing IDs from collections degrading the ecosystem?

Many argue that removing IDs from collections in response to hacks or bad actors is a major narrative violation about Bitcoin’s immutability (other ecosystems like ETH and Solana have structural enforcement on-chain which disincentivizes this, but Bitcoin doesn’t). Even some who have fallen victim to these actions will say that arbitrarily removing pieces from a collection isn’t worth it.

An important note: One thing to emphasize is that, even when someone decides to pull inscription IDs from a collection, the inscriptions are still on-chain and ordinal consensus still holds true. If you own the sat associated with an inscription (even a stolen one) the social consensus of ordinals ownership still enforces that. This is one strong argument in favor of actively managing marketplace collection recognizance. 

What is provenance anyway?

At a very high level, Bitcoin’s transaction chronology itself serves as a rudimentary provenance scheme. As Post Capone put it, “Even if parent child had never been added (it was a good add), provenance would still have been baked in.” Indeed, there are multiple ways to prove that a collection originated from a source or that a set of items are related. 

The only “ real provenance” enshrined in the Ord client, though, is parent-child and it can be unwieldy, expensive, and time consuming. Some believe that there are other, better ways to solve the provenance problem, but these solutions require time and advocacy to get off the ground.

“There is a level of elegance to the way parent-child works, and I see the appeal, especially for smaller collections.  There are also a number of significant tradeoffs of this model.  It’s not like other solutions don’t exist or can’t exist, they’ve just not been pushed for. I say this knowing I’m one such person who could have pushed for an alternative, but it has been hard to prioritize this.” – Nick Sainato, Gamma cofounder

That said, there have been many provenance schemes predating parent-child’s addition. Some of my favorite ones implemented are:

• Eloc’s Abstracti Art
• Timechain’s PGP key
• Bitcoin Burial’s exif data
• Onchain Monkey’s elaborate 10k inscription orchestration
• Bitcoin Shroom’s Timestamped SHA-256 fingerprints

Many ideas for on-chain provenance haven’t even been tried as far as I’m aware, such as Post Capone’s proposal to use BIP-322 from the transaction hash of an inscription.

My opinion and preferred path forwards

I do not like a future where the convention is to remove inscription IDs from a marketplace’s collection recognizance. Instead, I prefer some kind of flagging system on the marketplace side. I worry about things like personal grudges (creator doesn’t like holder X), the blurred line of user fault vs. an actual hack, and founder disputes (we all know how that went the other week). This also isn’t a condemnation against founders who have pulled IDs from marketplaces (e.g. I think that Trevor and Lifofifo are some of the best project leaders in the entire ecosystem).

My ideal scenario would be that we see emergent collection standardization. Imagine a convention on-chain that is publicly indexable by any third party and supported by inscription services and explorers. This is probably best done in the (woefully underutilized) metadata field. It would also probably require some kind of PGP key or BIP-322 signing from the founder to prove origination. This is hard to bootstrap and requires coordination and buy-in from many major stakeholders.

Which way, ordinal man?

I wrote this while jet lagged on a plane somewhere over the Pacific while on my way to Token 2049 in Singapore. I plan on hanging out with many of the folks involved in this provenance soul searching discussion and I’m optimistic we can find a way forward that doesn’t just incrementally improve upon “testnet” NFTs, but rather empowers them to become a long lasting testament to this exciting era of Bitcoin.