ERC-3475 and ERC-3525: A Dialogue

Three key issues at the heart of Web3 evolution topped a recent conversation between the creators of the ERC-3475 and ERC-3525 token standards.

Desmo Labs
5 min readNov 14, 2022

On October 14, 2022, D/Bond’s CEO, Yu Liu and the co-founder of Solv Protocol, Will Wang, held a roundtable discussion on the theme surrounding how their respective token standards — ERC-3475 and ERC 3525 — connect with each other with regards to Web3’s burgeoning decentralised web era.

Yiu and Wang had in-depth discussions and exchanges of views on the characteristics and design concepts of their standards including that both have multi-layered data structures hence fit for use to build advanced financial assets such as decentralised bonds. The discussion was also an avenue to shed light on the comparisons being made about the standards due to the overlap in their use cases.

  1. What a token represents to both protocols

They started with what a token means in the context of their respective protocols.

To Wang, at the core of the ERC-3525 token standard is its ability to act as a carrier of digital assets from the user’s point of view with two components at its core: the ownership and transfer relationships. He says the ERC-3525 standard has created a new ownership two-layer ID-to-ID transfer model.

ERC-3525 can be compatible with ERC-721 or ERC-5192. The standard can use its two-way architecture to build a two-layered asset model so that the protocol can have both ERC-20 and ERC-721 features. Wang believes that any digital asset model can be carved by the two dimensions of ID and slot.

From a global perspective, Liu describes the essence of a token to be a format for data storage and reading. Its underlying logic actually reflects a data structure, and the entire transfer is actually a complete read and write process. The ownership of the address corresponding to the token is the proof of the holder.

ERC-3475 has a set of data storage space that depends on the parameters in the data stored on-chain as metadata which is important for any protocol.

2. Importance of on-chain metadata

According to Wang, the design philosophy of a protocol should not be more complex as it may not necessarily be compatible with all scenarios. The core idea of ERC-3525 falls under the view that less is more or if it is not necessary, do not add entities. Therefore, based on this approach, they did not see a reason to integrate any on-chain data storage scheme while designing the protocol. He recommended that the original data of financial-related properties could be stored on chain or generated by the smart contract.

At the same time, he believes that they did not need to consider specific application scenarios for the design of the token protocol. Rather, they combined the characteristics of the protocol itself to bring it into a scenario suitable for the protocol. For example, the most significant feature of ERC-3525 is its dual-asset model. It can be reasonably applied in scenarios such as bonds or SBT.

Explaining the design idea of ERC-3475’s on-chain metadata storage, Liu described the token standard as a wheel.

“I could first see a problem that cannot be solved by existing solutions and then invent a wheel to solve this problem,” he said. “Or first invent the wheel, and then find a place where the wheel can be used.”

He believes that the outcome-determining process method and setting the purpose from the beginning will determine the trend and result. This is where Liu disagrees with Wang, or an indication of the divergence between ERC-3475 and ERC-3525.

3. Why Web3

Wang and Liu also tried to situate the ultimate goals of their initiatives within the overall Web3 ecosystem by seeking to identify what the main idea behind the new kind of Internet based on decentralised blockchains.

Since the concept of Web3 was proposed, there is still no clear definition, says Wang. He added that there is also the saying “100 people have 100 Web3” as he outlines how Web3 has conferred a wide range of topics from its birth to the present.

He argues that developers need not attempt to outline specific usage scenarios for a protocol, as this may greatly limit people’s imagination about how to use it. Rather, based on the notion that it is possible to provide solutions to a problem that need to be solved even before they arise, he said a protocol with certain features that coincide with the problem’s nature only needs to be brought into an application scenario.

Liu doesn’t think so. He argues that the essence of Web3 is to find a problem that Web2 cannot solve or could not solve well, and then provide an efficient solution for it. At the same time, this solution should have a certain degree of versatility so that it can still be solved even when encountering similar problems.

Other issues of relevance to D/Bond

Liu, who is into statistics and quantitative storage and has worked in a traditional financial company before going on to build a decentralised swap system for use by anyone regardless of their country or bank, says D/Bond is working on a solution with two scenarios: for Web2 and Web3-related companies. He said the idea is to provide them with a decentralised settlement and tokenisation process (including KYC protocols and all the supporting facilities of the white-label application) i.e. a system similar to Alipay in which any store can use it for settlement either on-chain or off-chain, centralised or decentralised.

There is also the multilayer pool technology which enables the use of a single instead of a dual currency pledge in D/Bond’s new automated market maker (AMM) system called automated pair maker APM. A bit like the Asset-Backed Security (ABS) in traditional finance where all the assets are tied together in one package, Liu says the technology’s benefits include reducing the risk of infrequent losses, slippage reduction, capital utilisation issues, and the bonding of assets in the pool.

He likened the early token standards (e.g. ERC-20, ERC-721) to a cell phone with a keyboard, and all of its storage format in physical form on the outside of the phone. Then the newer token standards like 1155, 3525 and D/Bond’s 3475 are more like a touchscreen phone. These new standards are constantly adding values even though their essence of existence is essentially different. In the case of ERC-3475, he said D/Bond is building a convenient platform that is strongly adaptable to and compatible with various use case scenarios.

In line with the vision for the future, Liu said D/Bond designed a set of data storage formats based on its abstracted storage standard that is easier to read and write on on- and off chain.


Liu concluded the interview by quoting a metaphor from a brief history of mankind. He compared Web2 to a horse-drawn carriage and Web3 to a diesel locomotive. With the use of locomotives becoming bigger on a large scale, the wagon days were doomed to be gone forever. He opines that it is why Web3 is needed.

The roundtable between ERC-3475 and ERC-3525 started with a focus on the token standards but gradually extended to core topics such as on-chain metadata and the essence of Web3. In the course of sharing their personal opinions, the two authors provided new ideas of looking at the same problem from different angles. The debate was a way for the teeming communities of both protocols to get closer to the essence of things.