Note that doing a PhD at ETH Zürich generally requires a Master's degree, but there is a <a href="https://inf.ethz.ch/doctorate/direct-doctorate-computer-science.html">direct doctorate program</a> that you can enter with a Bachelor's degree (application deadline December 15th).</i></p> -->
<p>My two main lines of work are about <a href="https://www.rust-lang.org/">Rust</a> and <a href="https://iris-project.org/">Iris</a>.<br>
-On the Rust side, I am working (also in collaboration with the Rust language team) towards a solid formal foundation for the language, including in particular the unsafe parts.
-One key result here is our <a href="https://plv.mpi-sws.org/rustbelt/popl18/">type safety proof</a>, which also describes a methodology for establishing type safety of well-encapsulated unsafe code.
-My goal is to make unsafe Rust just as safe as safe Rust by means of formal verification.<br>
-On the Iris side, besides continuing development of its logical foundations, I am interested in applying Iris to new problem domains; recently I started working on modular verification of fault-tolerant distributed system components.<br>
+On the Rust side, me and my group are working (also in collaboration with the Rust language team) towards a solid formal foundation for the language, including in particular the unsafe parts.
+As part of this we are developing <a href="https://github.com/rust-lang/miri/">Miri</a>, a practical tool for detecting Undefined Behavior bugs in unsafe Rust code, which has become a part of the standard toolbox of unsafe code authors.
+Meanwhile, <a href="https://github.com/minirust/minirust">MiniRust</a> is our work-in-progress proposal for a precise specification of unsafe Rust, that I hope to integrate into an official Rust specification eventually.
+My long-term goal is to make unsafe Rust just as safe as safe Rust by means of formal verification based on rigorous foundations for all key components of the language.<br>
+On the Iris side, I am continuing development of its logical foundations.
+We are making Iris fit for specifying and verifying programming languages at scale using a more modular approach.
+The long-term goal is for Iris to be able to handle the full scale of complexities that arise when doing foundational verification of real languages.<br>
For some more information, check out my <a href="https://www.ralfj.de/blog/categories/research.html">research blog</a>, my <a href="cv.pdf">CV</a>, and my <a href="research-statement.pdf">research statement</a>.</p>
<p>In my free time, I like to run internet services myself and work on free software.
projects / web.git / blobdiff