There has been much talk over the past two decades about commercialization of the mobile ad hoc network (MANET) technology. Potential ad hoc network applications with some commercial appeal are now finally emerging, “drafted” by the enormously successful wireless LAN technology. Closely coupled to commercial applications and critically dependent on commercial ad hoc networks will be the “pervasive computing” applications. Since military and civilian emergency MANETs have been around for over three decades, and since the Government has continuously supported MANET research for as many years, it may seem natural to assume that all the research has already been done and that commercial MANETs can be deployed by simply leveraging the military and civilian research results. Unfortunately, there is a catch. Commercial MANETs (and therefore pervasive computing applications) will evolve in a way totally different from their military counterparts. Most importantly, they will start small, and will initially be tethered to the Internet. They will be extremely cost-aware. They will also need to cater to a variety of different applications. This is in sharp contrast with the large scale, autonomous, special purpose and cost insensitive military networks. In this paper we review a typical “battlefield” MANET application and contrast it to two emerging commercial MANET scenarios—the urban vehicle grid and the Campus network. We compare characteristics and design goals and make the case for new research to help kick off commercial MANETs. In particular we argue that P2P technology will be critical in the early evolution of commercial MANETs and identify research directions for P2P MANETs.