Introduction to smart contracts

A smart contract is a user-supplied piece of code submitted to the Concordium blockchain, used to define behavior that is not directly part of the core protocol. Smart contracts are executed by nodes in the Concordium network according to predefined rules. Their execution is fully transparent, and all nodes must agree on what the outcome of execution is based on only publicly available information.

A smart contract can receive, hold and send GTU, it is able to observe some aspects of the chain, and maintain its own state. Smart contracts are always executed as a response to external actions, e.g., an account sending a message. In practice smart contracts will often be a small part of a larger system, combining on and off-chain functionality. An example of off-chain functionality could be a server that invokes the smart contract based on data from the real-world, such as prices of stocks, or weather information.

What are smart contracts for?

Smart contracts can reduce the needed amount of trust in third-parties, in some cases removing the need for a trusted third-party, in other cases reducing their capabilities and thus reducing the amount of trust needed in them.

Because smart contracts are executed completely transparently, in a way that anybody with access to a node can verify, they can be very useful for ensuring agreement between parties.

Auction smart contract example

A use case for smart contracts could be for holding an auction; here we program the smart contract to accept different bids from anyone and have it keep track of the highest bidder. When the auction is over, the smart contract sends the winner bid GTU to the seller and all other bids back. The seller then should send the item to the winner.

The smart contract replaces the main role of the auctioneer. The contract itself only governs the bidding part, and the on-chain distribution of GTUs. It will likely also need some logic for reimbursing the highest bidder if the seller does not fulfil their obligations. This will most likely mean that the contract needs to support some notion of proof that the seller has indeed fulfilled their obligation, or some way for the highest bidder to file a complaint. Smart contracts cannot resolve these real-world issues automatically, and the best solution is likely going to depend on the specifics of the auction.

What are smart contracts not for?

Smart contracts are a very exciting technology and people are still finding new ways to take advantage of them. However, there are some cases where smart contracts are not a good solution.

One of the key advantages of smart contracts are the trust in the code execution, and to achieve this a large number of nodes in the blockchain network have to execute the same code and ensure agreement of the result. Naturally, this becomes expensive compared to running the same code on one node in some cloud service.

In cases where a smart contract depends on heavy calculations, it might be possible moving this calculation out of the smart contract and have the smart contract execute only some key parts of the computation, using cryptographic techniques to ensure the other parts are executed correctly.

Finally, it is important to remember that smart contracts have no privacy and everything the smart contract has access to is accessible to everybody else on the Concordium network, meaning it is difficult to handle sensitive data in a smart contract. In some cases it might be possible to use cryptographic tools to not work with the data directly, but rather have the smart contracts work with derived notions such as encryptions and commitments, which hide the actual data.

Life cycle of a smart contract

A smart contract is first deployed to the chain as part of a contract module. After this a smart contract can be initialized to obtain a smart contract instance. Finally a smart contract instance can be repeatedly updated according to its own logic.