This is the first in a series of posts laying out what we have learned in the chain search and concepts that are needed to understand where we are going. Up first? A look back at the problems we’ve encountered and how our needs were mismatched with what was available.
In 2020, we released the Decentralized Social Networking Protocol (DSNP) Whitepaper. Since then we have been working to make that protocol a reality. We started with a proof of concept on Ethereum with version 1.0 of the specification. Although Ethereum has issues such as cost and possible transactions per second (TPS), it is a great ecosystem of developers, knowledge, and tools. We used Ethereum as a testing ground knowing that we would need to move to a different chain eventually.
Along the way we refined the ideas from the whitepaper and tried to focus on the reality of what is possible. Several times we questioned the core reasons we wanted to use the blockchain in the first place. Each time finding alignment in our principle of having the users control their own data and identity. Users should be the ultimate power holders in this new ecosystem.
In addition to being the shiny new toy in the bin, blockchains provide real value through a system enabling user ownership, choice, provider cooperation, and collaboration in ways previously difficult or impossible. Achieving user ownership in technology has historically required highly technical participation, if allowed at all. Anyone can, in theory, physically run a server, but in practice that means technically inclined individuals of means. We want to follow the trend of expanding participation and access, not reverse it. Blockchains provide a foundation for ownership without the need for equipment.
Ownership and Choice
Ownership is fundamentally about the ability to choose. Choose who you want to interact with and who you trust. When users are the product and only have the power to leave at the cost of forfeiting all they have built on a platform—trust is replaced with lock-in. Blockchain elevates and expands the options for participation in how the network functions and provides for innovation beyond walled off APIs. In addition, we gain benefits from the systems of cryptography, decentralization, economic incentives, and more.
These qualities encouraged us to find a blockchain that could be the foundation for DSNP. We looked at dozens of chains. We had various evaluation criteria to try to weigh the different tradeoffs, but three interconnected issues were above all the others: Low cost, scale, and decentralized disinterested nodes.
Low Cost
Low cost is a bit of a holy grail for blockchain. Every new chain claims advances in transaction price reduction. However, as the chain becomes popular the transaction cost in fiat terms will skyrocket. While some of the value is speculation, a portion is also due to the increasing use and recognition of the usefulness of blockchain. That same volatile rise in value leads to unpredictable costs which limit the ability for many services to build on top of blockchain, especially at scale. How can we convince users and services to adopt the DSNP if the cost of using the protocol fluctuates from next to nothing to tens of dollars per chain transaction?
Scale
Scaled blockchains have been making strides with some chains passing hundreds of thousands of transactions per second. Yet, the scale of social media interactions when you include posts, replies, reactions, direct messages, and more is orders of magnitude greater. DSNP leverages a batching concept for compressing the vast number of these messages from a protocol standpoint, but the scale is still beyond what many blockchains are able to do in addition to their existing growth and load.
Decentralized Nodes
Scale aside, some blockchains we looked at have reduced cost by increasing centralization. While we expect chains to go through a period of centralization during testing and early adoption, disinterested nodes are critical for maintaining user ownership when forces collude. For DSNP to remain an open protocol, decentralization of the infrastructure is key to provide guarantees against powerful bad actors. Social media is targeted by a wide variety of actors who desire to control the flow of information. China’s “Great Firewall” is the best known, but most political powers have tried with various degrees of success to control the flow of information for good and ill. In the end, infrastructure is not a moral agent and cannot not be required to make moral choices. Infrastructure should enable moral decisions to be made by users, service providers, and democratic governance. (Look for more on DSNP governance in a future post.)
Trade-offs
Cost, scale, and decentralization are the largest requirements for a blockchain to support DSNP. We also found a secondary tier of competing trade-offs while investigating potential chains. Each blockchain comes with its unique spectrum of solutions to the problems facing all blockchain projects.
- What is the transaction load that the chain can support?
- What happens when the chain is under congestion?
- How can we support shifting network costs from users to services without losing user ownership?
- What are the data storage and availability guarantees?
- How can we support upgrades and improvements over time?
- What are the protections against spam and network abuse?
- Is the economic system of incentives able to support growth, contraction, and steady state situations?
- Is the development of the project secure to continue?
- What outside incentives and pressures might the project have?
- How hard is it for developers to onboard to build a larger ecosystem?
The Perfect Chain
As we tried to answer these questions and more, something kept bothering us. We felt stuck in a blockchain version of Goldilocks and the Three Bears, except with a lot more bears and no bed that matched our needs. Some chains appeared great, but then as we looked closer, the price might shoot out of our constraints. Others proclaimed decentralization, but were actually dangerously controlled by a few parties.
During the process, we realized that almost every chain we looked at was focused on the value of each individual transaction. In DSNP, as with networking in general, the value is not in the individual messages, but in the aggregate of a scaled number of messages and the possibilities of potential connections that creates value.
Another realization was that our messages on DSNP have different constraints than financial transactions. For DSNP, as other networking protocols like TCP/IP, the messages are being created not value transferred. For a new blog post, no balance needs to be validated before it can be created. An act of creation is fundamentally different from an act of transfer.
As we understood these constraints and tradeoffs, an outline emerged of a chain perfectly designed for DSNP (or other protocols with similar needs) could look like:
- Decentralized disinterested nodes
- Scaled messaging with origin validation
- Predictable costs for users and services
- Extremely low cost on a per message basis with either support for coinless transactions for users or cost shifting from users to intermediary service providers
The world is different from 2013 when Vitalik Buterin expanded on the ideas proven at scale in Bitcoin. We stand on the work of countless developers today. We believe we can solve the next set of problems to build a blockchain that provides scalable message passing at minimal cost. A chain that can be a backbone of new shared protocols. A world that expands on the ideals of the Internet to create value through cooperation and user choice.
Next: Concept Series 02: Source-Dependent Messages
Edits:
2022-04-10 Added headers for easier readability.
2022-04-20 Added link to next post