Sikoba — Update #27
TLDR: focusing on mainnet launch
We are now focused entirely on the launch of the Sikoba mainnet. Updates will be published ad hoc, instead of on a monthly basis.
In April, Razvan Rosie has joined the team as Chief Scientist. Before joining Siokba, Razvan was a researcher at Luxembourg University’s Applied Crypto Group for two and a half years. His experience also includes several years of work as software developer for commercial projects. Razvan studied Computer Science at Imperial College, London, obtained his PhD at Ecole Normale Supérieure, Paris and has published numerous scientific papers, mostly on encryption primitives.
Since our last update, Crystal 1.0 has been released, which is a really good thing given that we are going into production soon. We remain convinced that Crystal is a good choice for this project, as it combines elegance and speed.
We have updated the ECDSA.cr shard to be compatible with Crystal 1.0, also Razvan added a native Crystal SHA3–256 implementation which is very fast. Caching has been added to improve the performance of ECDSA signing and verification, although we’re still a long way from the highly optimised performance of OpenSSL.
In mid-March, Alex published “An Introduction to Asynchronous Binary Byzantine Consensus” which shows, as a very simple corollary, that asynchronous binary consensus is possible in weakly Byzantine environments. Combined with previously published papers on “Proof-of-Decision: Achieving Fast and Timeout-Resistant Consensus in Asynchronous Byzantine Environments” (September 2019) and “One-Step Consensus in Weakly Byzantine Environments” (March 2019), this opens the way to creating a very fast consensus layer for the Sikoba blockchain in the future, although we will of course initially use Babble as planned.