Thursday, 10 November 2016, 14:00
Cybernetica Bldg (Akadeemia tee 21), room B101
Abstract: The goal of incremental computing is to efficiently update a view of some data source whenever the data source changes. Our previous approach to incremental computing was based on change propagation, where changes were represented by values of algebraic data types. In this talk, I present a new way to compute incrementally. Its central idea is to not work with changes at all, but to use the evolution of a data item as the core concept. This new approach to incremental computing is more functional and hopefully more powerful.
This is joint work with Denis Firsov.