Sharding Is Already Ushering in Radical New Ethereum Designs
So-called "sharding" may still be theoretical, but the promising implications of the concept are becoming more and more real.
At least that's the case on†ethereum, where developers are beginning to see the scaling solution, which would essentially split the blockchain into parts that would run on different servers, as an opportunity to test fundamental assumptions about one of the world's largest cryptocurrencies.
Although initial roadmaps are just now being discussed, ambitious coders†are already jumping to introduce protocol-level redesigns that could be made possible by the upgrade.
"Sharding is a huge, huge change to the network," said Phil Daian, a researcher at Cornell University's Initiative for Cryptocurrency and Contracts (IC3).†"A lot of people think it provides an opportunity to redesign economic models and other aspects of the system."
For Daian, the†realization comes on the heels of a developer retreat in Taipei, where, along with an all-star team of co-founders including Ari Juels,†Lorenz Breidenbach and Florian Tramer, he discussed a new initiative aimed to redesign ethereum to work more efficiently,†Project Chicago.
The project is trying to identify exactly what commodities are being traded at the core of ethereum today. By isolating†a variety of network elements, like its gas, storage and UTXO transaction data, the team plans to implement protocol-level markets for what they call "crypto commodities."
"We want to look at all of the services and resources the network is providing and say, 'OK, how do we create a market-based system for price discovery and the incentivization of this,'" Daian told CoinDesk in interview.
The researchers were inspired to create the concept after developing a tool called†GasToken, which allows ethereum users to store gas (ethereum's token for paying fees on the network) when it's cheap and sell it at a later date when the price is higher. To do this, users must send a deliberately faulty transaction that doesn't go through the network and, hence, are able to reclaim the gas through a refund mechanism.
And while not many people are using the tool yet, it's effectively shone a light†on an incentive flaw within the ethereum system in that, as people look to store gas, it further bogs down the ethereum state - the part of the system that keeps track of all possible computations.
Already, the incentive flaw is†reigniting discussions†about the need for users to pay so-called "rent" on the amount of time they need their data to be stored on the blockchain. But because GasToken incentivizes people to hoard gas, "it's a clear artifact to point to show people why today's model is flawed and why rent needs to be introduced," Daian said.
Still, this isn't the only thing the researchers at Project Chicago think needs to be redesigned.
And as such, Daian spoke more broadly about sharding, stating:
"It could actually provide a once-in-a-lifetime opportunity to radically redesign the system and reset people's expectations from scratch."
Futures market inspiration
That's because, according to Project Chicago, at its core, a blockchain is a marketplace, one where miners sell resources allowed by the software to users. Focusing on this, Daian last week drafted an incentive scheme for peer-to-peer networks, one that would not only pay participants for routing transactions, but apply the same logic elsewhere.
"These resources can be anything from block space, to CPU on full nodes, to permanent storage on full nodes, etc. So, we sort of came at this from the beginning, questioning the pricing models that blockchains have today," Daian said.
Created earlier this year, GasToken was the first step in this direction. Exploiting a feature named "gas refund," the contract allows users to store gas when it's cheap and resell it when it becomes more expensive.
Daian described this as a "fundamental mis-pricing" in ethereum, in that it values computation as equivalent to storage. "Because of that, we've now created a direct financial incentive for people to bloat the state space and store garbage," he said.
As well as revealing inefficiencies in ethereum's incentive structure, GasToken paved the way for a line of inquiry that could be extended deeper into the protocol layer.
"It sort of made us realize that there's this whole under-researched space of how to deal with these these raw resources that are fundamental to different blockchains today," Project Chicago's Tramer told CoinDesk.
By identifying markets for raw resources, Project Chicago intends to pave the way for other financial mechanisms, such as futures. "[We'll be] looking at different kinds of futures for ethereum, computation, storage and network, and how you can build them," Daian said.
According to Tramer, by speculating on the availability or scarcity of the underlying resources over time, such markets could potentially mitigate price volatility, just like on traditional markets.
Daian echoed this, telling CoinDesk:
"There are really concrete analogs to the real world here. The Chicago Mercantile Exchange (CME) was our inspiration for Project Chicago. And I think a lot of real-world problems could have been avoided by a nicer economic model."
However, Daian is aware that by ramping up the markets, such schemes may not prove popular.
For example, an increased number of incentives could lead to centralization, attracting large-scale players to participate in storing or mining the blockchain in exchange for rewards. Daian deflected this though, stating, "My argument would be that you're essentially saying you've introduced an incentive and now it will be vulnerable to economies of scale."
He continued to say that bigger economies are both positive for security, in increasing the cost of attacks, and an inevitable economic progression, "even if you do fight them,"referring to monero's recent efforts to defend against large-scale mining.
But there are other potential issues to the mindset, as well. While Project Chicago could provide incentivizes for a host of new participants, such schemes would come at a cost.
For example, rent would mean that token issuers pay a yearly fee to host a smart contract on ethereum, which, failing renewal, could lead the contract to be deleted.
According to Daian, it's possible that new charges could drive users away. "It is worth saying that for all of these crypto commodities, a big is risk in my mind is that people sort of like the cheaper, subsidized model," he said.
Plus, in a competitive market for blockchains, new cryptocurrencies could emerge that offer free usage in the short term, "because there's not that much demand, and perhaps there's good supply."
And while the new incentives could be a big improvement for speed, as well as scaling and decentralization, it's not clear how much those attributes are valued by users.
"I think people are going to have to start valuing decentralization and incentive compatibility and robustness with their wallets. And it's not really clear to me whether or not they do."
Shards of glass†image via Shutterstock