Fwd: [security-lunch] Oct 29 | Dale Schumacher on "Security All The Way Down"

0 views
Skip to first unread message

Alan Karp

unread,
Oct 27, 2025, 6:28:41 PM (2 days ago) Oct 27
to <friam@googlegroups.com>


--------------
Alan Karp


---------- Forwarded message ---------
From: Michael Leo Paper <mpa...@stanford.edu>
Date: Mon, Oct 27, 2025 at 3:27 PM
Subject: [security-lunch] Oct 29 | Dale Schumacher on "Security All The Way Down"
To: securit...@lists.stanford.edu <securit...@lists.stanford.edu>


Security Lunch 🍂 Ed. — Wednesday,  Oct 29th, 2025, 12:00 pm @ CoDa E160

Security All The Way Down
Dale Schumacher
Can't make it in person? Join us on zoom.
See our past & upcoming events on our website

Abstract: 
Two of the most significant architectural factors in computer security are interference through shared mutable memory and temporal coupling through synchronous call/return interaction. The µFork processor architecture addresses both of these issues at the machine level. Processing is driven entirely by asynchronous events. There are no blocking operations. Instruction execution is efficiently interleaved, so all ongoing activities make progress concurrently. Activities are strongly isolated from each other. State-change is always local. Memory safety and resource quotas are enforced in the machine-language. Interaction occurs by generating asynchronous events. This simple design uses less logic and consumes less power to maintain coherency and prevent conflicts. µFork security can be summarized as, “If you don't have it, you can't use it.” You have to know an address to generate an event for that address. The address-reference graph is the access-control graph. This architecture has been implemented in a Rust/WASM component that may be embedded in a web page or run in a JavaScript container. A binary-compatible FPGA implementation is under development.

Bio:
Dale Schumacher has been working with computers professionally for over 40 years. The last 15 years he has focused primarily on the Actor Model of computation; exploring implementation strategies, usage patterns, security implications, and performance characteristics. He has implemented several actor runtimes and languages, including his own pure-actor language Humus. He collaborated on Carl Hewitt's ActorScript language. His focus remains on the practical application of actor-oriented technology for secure fine-grained concurrent computation with minimal resource usage.
_______________________________________________
security-lunch mailing list
securit...@lists.stanford.edu
https://mailman.stanford.edu/mailman/listinfo/security-lunch
Reply all
Reply to author
Forward
0 new messages