t.co

has an advanced software specification language supporting knowledge representation. People write about real-life ideas in human readable language; how the software should function. Tau's logic engine then outputs provably correct software.

August's Development Video Update is Live! 🎥 https://youtu.be/x9R-8NJGG20👈 Summary: 🔹Ohad's been working with Prof. Franconi and Dr. Paweł Parys on the new Tau language. It has the temporal software specification part, which they got to understand better. It's basically done and very promising! They mainly continued with the knowledge representation part. The goal is to translate knowledge representation languages, also called description Logics, into our existing engines, which are Boolean algebra engines. Doing so means the whole thing will be much less complex and much faster developed. On this front, they've made significant progress in showing how very rich description Logic encodes into Boolean algebra. Why hasn't this been done by someone before? One would need to know about good decision procedures for Boolean algebra, which did not exist until Ohad found a solution. Regarding the Tau language, Ohad's managed to answer all open questions regarding the software specification and almost all available questions regarding description Logic, and knowledge representation. Ohad's beginning to implement the Tau language so people can play with it ASAP. 🔹Juan, who's been focusing on first-order Logic, has started working on an implementation that's a transformation initially suggested by Ohad, which takes first-order logic programs and turns them into disjunctive normal formulas or plain TML formulas for execution. A new approach for handling first-order Logic in TML. He's currently finishing up plugging in that transformation algorithm into the TML runtime. 🔹Tomáš undertook a significant task of splitting the software BDD library and the Earley parser, developed this year, into two libraries and repositories that will promote reuse in future projects. He's working on another task from Ohad; removing dependence on raw program structures and going directly from parse trees to Cloud programs. A task that everybody has been working on this month. 🔹David has finished his work on squaring of TML programs and is currently performance testing. He's also re-implemented the query containment using the Z3 engine and changed a lot of code, eliminating raw programs going directly to flat programs. He's reworked query optimizations, written unit testing for all the individual functionality, and tested everything using randomly generated TML programs. He's using a static analysis tool called @SonarLint to detect issues with code and to improve code quality. 🔹Andrei has worked with our tester to fix issues uncovered in the first two testing phases. He's continuously improved the core of the application and the areas of video conferencing and scheduling. He's trying to make it easier for users to acquire Agoras to pay for video sessions. He's doing that by connecting directly to distributed crypto exchanges. 🔹Luis, on Logic and decidability, has been working on finite tree automata and tree languages and guarded fixed point logic with tableau. He's working on weekly technical articles on central topics of logical Ai.