Deploying a computationalmarket for resource alloca- tion in the systems domain can benefit two research con- stituencies. The first constituency, which will be ignored for the rest of this paper, are the experimental economists and economically-minded computer scientists. Rarely are economists actually given the opportunity to deploy a market or a whole economy, let alone several for compar- ison. Computationalmechanismdesign [13] is an emerg- ing topic partly because the results apply to many differ- ent domains, and there is some merit in asking systems researchers to be research subjects as they attempt to use some market mechanism for their own work. But systems researchers (the second constituency) are much more interested in knowing.
