[E-Lang] New Page: Partially Ordered Message Delivery
Mon, 05 Mar 2001 18:44:02 +0000
"Mark S. Miller" wrote:
> At 09:10 AM Monday 3/5/01, Ben Laurie wrote:
> >Tyler Close wrote:
> >> I think E's archash technique is actually more flexible than this. In
> >> an E message send, the sender gets to specify the address that the
> >> return value will be placed at. The sender does this by providing a
> >> nonce (the archash) that will be transformed by the receiver with a
> >> one-way-hash (SHA1), to calculate the return value address.
> >Hmm! How is this calculation done?
> The archash is a randomly chosen unguessable number. Or actually,
> pseudo-randomly chosen from an entropy pool into which we're constantly
> tossing new alleged entropy whenever we find some (Entropy mechanism thanks
> to Bill). The "address" (as in SwissNumber) is simply the SHA1 hash of this
> unguessable number.
> I called it "archash" as an abuse of trig terminology.
> X == archash(address) iff hash(X) == address
Ah, OK. And what's the point of using the hash instead of the address?
"There is no limit to what a man can do or how far he can go if he
doesn't mind who gets the credit." - Robert Woodruff
ApacheCon 2001! http://ApacheCon.com/