Inside TrueBit: Ethereum's Lesser-Known Scalability Effort
An under-the-radar effort to supercharge ethereum smart contracts is gaining momentum.
While it may seem like another in-progress scalability project, TrueBit is distinguished by its developer team, including developer Christian Reitwiessner, creator of the network's Solidity smart contracting language, and mathematician Jason Teutsch.
Announced last year, TrueBit's goal is to enable support for more powerful smart contract computations on the distributed application platform – the most ambitious of which are those that would be needed for applications like rendering photos or machine learning and artificial intelligence.
The scaling issues inherent in public blockchain architecture are known, but computational power could prove a significant problem for ethereum, which looks to be a 'world computer' that supports a richer variety of complex applications similar to those found in an average app store.
The creators foresee a way to increase ethereum's computational power by harnessing another distributed network that would conduct and verify computation, while disagreements would be settled on the ethereum blockchain. In theory, this would limit the workload placed on the lower levels of the system, where data is stored by a large network of global nodes.
TrueBit founder and University of Alabama at Birmingham postdoc Jason Teutsch told CoinDesk:
”[Today,] smart contracts can only do very trivial tasks from a computational point of view. Basically, what TrueBit does is to give smart contracts the ability of doing scalable computations.”
So, what's wrong with the system today? Currently, ethereum has a 'gas limit' that puts a cap on the network's computational power per block.
This is roughly equivalent to bitcoin's limit on the transactions it includes in each block, though the gas limit is dynamically set by miners as opposed to being hard-coded into the network.
To date, the gas limit has already given rise to issues. For example, miners last year temporarily lowered the gas limit to thwart attacks on the network, impacting the operations of ethereum apps and companies.
Yet, without such a limit, Teutsch said, a problem known as the 'verifier's dilemma' arises, and miners are incentivized to accept unverified scripts into the blocks they mine.
"If you allow unbounded computation with ethereum smart contracts – the way they are right now – you would not only get denial of service attacks, but you would also end up getting wrong answers on the blockchain," he said.
Finding a workaround to these limits is where TrueBit and similar projects come in.
Like other next-generation blockchain projects, TrueBit uses a layer above the blockchain to do the heavy lifting. In this case, it outsources the verification of computations.
Rather than each and every node computing each and every smart contract, participants in the market – potentially anyone who owns a computer – perform this task. These participants are called 'solvers' and they submit a solution to the problem for a reward, while 'validators' check their work.
The project's white paper describes the TrueBit system as a 'verification game', where a market of off-blockchain computers verify computations. If even one participant disagrees with a solver's result, they can kick it to the blockchain to settle the dispute.
And the system aims to incentivize players to act in a trustworthy manner.
"At the end of this game, either the cheating solver will be discovered and punished, or the challenger will pay for the resources consumed by the false alarm," the paper explains.
Already, the computationally heavy distributed application Golem, a market for CPU and GPU power, intends to use TrueBit, as opposed to the ethereum blockchain.
However, there are other applications for TrueBit, Teutsch suggested. Dogethereum, a project aiming to connect the dogecoin blockchain to ethereum's, is one he is particularly excited about.
Dogethereum is different from other 'bridge' attempts, such as BTC Relay (which bridges bitcoin and ethereum), in that it could allow users to export currency instead of having to switch currencies between blockchains.
BTC Relay allows users to pay for ethereum applications with bitcoin, but it does not transfer the currency to ethereum, Teutsch said.
"In contrast, the Dogethereum bridge, as we envision it implementing with TrueBit, would effectively allow users to export currency from dogecoin to ethereum without (necessarily) making dogecoin transactions visible to ethereum smart contracts," he said.
Others have come to TrueBit with application ideas, as well, such as using it to stream video on Livepeer, which Teutsch described as a "YouTube without the YouTube". And still others, such as Gnosis, are working on separate off-chain computational scaling.
Will TrueBit help power this new wave of applications?
That answer seems uncertain. However, Teutsch said that it looks like they might deploy the technology on both ethereum classic and bitcoin-bound smart contract platform Rootstock in the near future.
Either way, the project shows the continued demand for scaling, and the many ways developers on the ethereum network are taking to attack the issue.
Light and shadows image via Shutterstock
Maine Lawmakers Kill Blockchain Voting Study Proposal
Zcash Releases Software Fix After Denial-of-Service Bug...