grant matcher

[Raoul Duke]

http://www.erights.org/elib/equality/grant-matcher/

i think since the graph is fixed, there's formally no solution. (even
with eq who says i trust the GM to not be buggy or malicious?)

'once you fix the capability graph and the introduction rule, a bunch
of “obvious” security impossibilities become literal graph‑theoretic
theorems.'

so i would want capability systems to implement checkers that use
formalizations to tell users what is/not possible?

[Mark S. Miller]

On Sun, May 3, 2026 at 6:05 PM Raoul Duke <rao...@gmail.com> wrote:

http://www.erights.org/elib/equality/grant-matcher/

i think since the graph is fixed, there's formally no solution. (even
with eq who says i trust the GM to not be buggy or malicious?)

What? I don't understand. I claim that E did solve both local and distributed EQ, and building on that, GM is trivial (which is the point).

 

'once you fix the capability graph and the introduction rule, a bunch
of “obvious” security impossibilities become literal graph‑theoretic
theorems.'

so i would want capability systems to implement checkers that use
formalizations to tell users what is/not possible?

--
You received this message because you are subscribed to the Google Groups "cap-talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cap-talk+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cap-talk/CAJ7XQb7V-hZFYqiUix64UWyd%2BddhetbSaOp9mg-ahRxXsN0-Jw%40mail.gmail.com.

--

  Cheers,
  --MarkM

[Raoul Duke]

Sorry, I was unclear & behind the times, on the old "without eq" variant.