How do you think of this abstract of the withdrawn analysis paper

https://eprint.iacr.org/2019/864
" Another Look at Byzantine Fault Tolerance

Yongge Wang

Abstract: We review several solutions for the Byzantine Fault Tolerance (BFT) problem and discuss some aspects that are frequently overlooked by existing literatures. For example, PBFT and HotStuff BFT protocols (HotStuff has been adopted by Facebook Libra) require a reliable broadcast primitive. We show that if the broadcast primitive is not reliable then the PBFT and HotStuff BFT protocols could not achieve the liveness property (that is, the system will never reach an agreement on a proposal). Though these BFT protocols have been developed for partial synchronous networks, we show that they cannot achieve consensus in partial synchronous networks since the participants do not know what is the Global Stabilization Time (GST) and broadcast channels before GST are defined to be unreliable (e.g., DoS attacks on certain participants). Thus it is important for developers to be aware of these issues when developing applications (such as blockchains) using these BFT protocols."

3 Likes

Our team looks forward to working with the author once the withdrawn paper is resubmitted.