ProfHewitt

ActorsTheory serves as the rigorous mathematical foundations of capability systems.

Foundation is about rigorously specifying the general Laws of Locality for Actor systems that CHERI implements.

Great to see further progress on CHERI!

See the following for theoretical foundations for the work:

Linux 60th Anniversary Keynote

https://t.co/IRe3vpMlWn

ProfHewitt

5 年前議論

Interesting article.

Mathematical foundations are needed that are much more suitable to modern Computer Science than using "universes".

For an example see the following article:

https://papers.ssrn.com/abstract=3603021

Also, see the following video:

https://www.youtube.com/watch?v=AJP1VL7shiI

ProfHewitt

5 年前議論

Computer science theory didn't get harder.

Instead, education became weaker :-(

See following link for foundations of computer Science:

https://www.youtube.com/watch?v=AJP1VL7shiI

ProfHewitt

5 年前議論

The Actors Abstraction does not require specification of

computer boundaries.

ProfHewitt

5 年前議論

The Actors Abstraction is foundational for computer science

analogous to the Natural Numbers in classical mathematics.

ProfHewitt

5 年前議論

Probably not a good idea to burden application programmers

with having to annotate each message send with whether the

receiver is on the current computer or another one in a

Citadel.

Since Actors can move between computers in a Citadel, such

annotations are infeasible in practice.

Of course, modularity, security, and performance should not be

ignored.

ProfHewitt

5 年前議論

See the following for a rigorous definition of the Actors Abstraction:

https://papers.ssrn.com/abstract=3418003

ProfHewitt

5 年前議論

Probably can't make much progress without a scientific foundation including the Actors Abstraction.

ProfHewitt

5 年前議論

The Actor Paradigm is much more widespread with offerings from Microsoft, Lightbend, etc.

ProfHewitt

5 年前議論

I fondly remember Joe Armstrong, who was a good friend.

However, not having a mathematical foundation has been a

severe limitation for Erlang.

The Erlang community is trying to dig its way out with

modifications and extensions. Still have a long way to go.

Of course, no Actors framework is perfect.

Many important implementations have been done Erlang.

Also, significant Erlang projects are ongoing to good effect.

ProfHewitt

5 年前議論

Erlang has absolutely no security :-(

ProfHewitt

5 年前議論

In a Citadel, distributed Actors are going to be necessary.

See the following:

    Citadels: faster response time and better information integration than remote datacenters

https://papers.ssrn.com/abstract=2836282

ProfHewitt

5 年前議論

Also see offerings from Lightbend and Microsoft.

ProfHewitt

5 年前議論

Erlang incorporated some of the ideas of the Actor Abstraction

but missed a few things. See

See article and video here:

https://papers.ssrn.com/abstract=3603021

https://www.youtube.com/watch?v=AJP1VL7shiI

ProfHewitt

5 年前議論

Swift Actors are an important contribution to the Actor Paradigm.

There is is an upcoming keynote address on the Actor Paradigm at the Linux 30th Anniversary Conference. See

https://reactivesummit2021.sched.com

ProfHewitt

5 年前議論

Both ASIO and Sender/Receiver are not a suitable foundation for concurrency in C++.

ProfHewitt

5 年前議論

As dueling co-routine proposals neither ASIO nor Sender/Receiver are a suitable foundation for C++ semantics.

ProfHewitt

5 年前議論

Async co-routines are not a suitable foundation for C++ semantics.

ProfHewitt

投稿

Apple Adopts Actors

Video Rembrance of Marvin Minsky (part 1)

Church's Dilemma Resolved

Gödel failed to prove inferential undecidablity in foundations

The Church/Turing Thesis Is False

Hundreds of Times Faster Than a Parallel Lambda Expression

Further Development of Dijkstra's Contributions

Rescuing Foundations of Computation

Edsger Dijkstra Tribute

Physical Indeterminacy in Digital Computation

コメント