Martin Geisler wrote: > [EMAIL PROTECTED] writes: > <cut> > >> Are there really other problems than sharing (and opening a shared >> value towards only one party)? > > I don't think so. All other operations (addition, multiplication, > comparison, etc...) are done on variables which represent values in > the ideal functionality implemented by VIFF. So they are not tied to > any particular player and so I think we can limit these asymmetric > operations to sharing and opening. >
I agree only sharing and opening will be asymmetric, and there are many cases in which an opening will be asymmetric. To make it really confusing there might be cases in which the opening might be dependent upon some secret shared variable. E.g. lets say you are in a hospital and that the nurses (doctors and administrators also) would have Pda's with location information. Now the nurses do not want their location to be tracked, so there is no central location server. Now one nurse puts out a request for help, and the algorithm has to find the closest nurse that is not busy to give the request to. The calculations to find the closest nurse is done using MPC, but then the call for request should be revealed only to the one fitting the description. The others should just get a dummy request (logged or discarded by the computer). In practice the nurse who's assistant was needed would hear a ring tone, and the pda would display that help was needed in room XXX. For all the other nurses (doctors and administrators), their pda would only log that an event where location information was needed was processed by the MPC module. They would not get any information about what kind of request it was, or anything else. In this example who the receiver of the message is, is not known until we have determined who is the closest. As a side note: I am working on a new comparison protocol, I have some problems with it so it might not work, but if it does then I need to compute both in some field Z_p and Z_q for different primes p and q. Can VIFF handle that or is it hard wired to only work in one field? _______________________________________________ viff-devel mailing list (http://viff.dk/) [email protected] http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk
