On Friday, March 25, 2016 1:33:16 PM CEST Andrew Gallagher wrote: > Before we even *think* about a protocol, there are policy hurdles to be > overcome, e.g.: > > 1. What criteria should be met before a key is removed?
Owner of private key or owner of UID/email address requests it. > 2. Who decides that the criteria have been met? The keyserver operator the request is sent to. > 3. How are malicious removals prevented? If owner of private key and owner of UID/email address disagree, the key stays off the servers. If they agree there should be no malicious removal. > 4. How is whack-a-mole prevented? It is not prevented, but right now you wouldn't even have to play it. > These are all *hard* problems, and none of them have much, or anything, > to do with protocol design. I don't see how they are. I see _all_ the hardness in the protocol design and implementation. Sincerely, Malte -- 1BEA 8159 A070 2E53 0152 A59F 0CC5 76E9 703E 1DDC _______________________________________________ Sks-devel mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/sks-devel
