[ntp:questions] Re: server's address in ntp payload?

David Schwartz davids at webmaster.com
Mon Nov 21 22:04:38 UTC 2005

"David L. Mills" <mills at udel.edu> wrote in message 
news:dlnfck$kbd$1 at dewey.udel.edu...

> The server private value need not and is not divulged. It is hidden by the 
> hash operation, just as in very many other protocols. I don't know why we 
> are having this discussion, as the issues at hand are common with other 
> secret schemes. See the Autokey briefings on protocols and identity 
> schemes at www.eecis.udel.edu/~mills/ntp.html.

    Perhaps you don't recall how we got into this discussion. We were 
talking about whether or not the server's address belongs in the UDP data 
payload. My position was that it belonged there. You said that the server's 
private seed could serve the same purpose the address would serve. I argued 
that it would not.

    Somehow you concluded based on this that I had some issue with the way 
the hash was managed. I have no such issue. It is perfectly designed to 
serve its intended purpose. However, it does not obviate the need for a 
common unique identifier for the server, akin to a public key.


More information about the questions mailing list