We know this -- that's how the system currently works.
Well, let me amend this -- we know you can do this so that it takes the spammer/attacker some non-trivial amount of time to break your measure (i.e., do a blackbox analysis on your new algorithm and defeat it) such that you already have a fix in place, or can have a fix in place quickly enough that the attacker doesn't reduce your utility below that of your competitors'.
The existing system works "well enough". Would an open system work as well, worse, not at all, better, or perfectly? Those are going to be hard questions to answer empirically, and harder questions to answer with proofs.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(no subject)
Date: 2008-01-04 06:56 pm (UTC)Well, let me amend this -- we know you can do this so that it takes the spammer/attacker some non-trivial amount of time to break your measure (i.e., do a blackbox analysis on your new algorithm and defeat it) such that you already have a fix in place, or can have a fix in place quickly enough that the attacker doesn't reduce your utility below that of your competitors'.
The existing system works "well enough". Would an open system work as well, worse, not at all, better, or perfectly? Those are going to be hard questions to answer empirically, and harder questions to answer with proofs.