Taming Smart Contracts With Blockchain Transaction Primitives: A Possibility?

Decentralized applications (DApps) rely heavily on user-programmed transaction protocols called “smart contracts” as fundamental building blocks for implementation. Such user-programmed transactions become necessary due to the limited out-of-the-box transaction types typically offered natively by blockchain platforms. While smart contracts offer significant flexibility and customizability in implementing transaction be-havior, they also come with significant usability and performance limitations that may impede the adoption of blockchains in specific application contexts. Motivated by the success enjoyed by database systems due to the systemization of a reasonable set of system-supported primitives, we consider the possibility of a similar approach for blockchains. Specifically, we utilize a combination of manual and automatic methods to analyze 93GB of transaction data from the Ethereumblockchain to understand whether there are transaction patterns that emerge and can be considered candidates for such automation within blockchains. We provide an overview of the implementation strategy for such transactions. The results of our analysis provide strong indicators in support of this approach.