The lack of HCI provisions in the invention of novel software applications and the bias of design knowledge towards desktop GUI make it difficult for us to design for novel scenarios and applications that leverage emerging computational technologies. These include new media platforms such as mobiles, interactive TV, tabletops and large multi-touch walls on which many of our future applications will operate. In this presentation I will discuss that novel application design should come not from user-centred requirements engineering as in developing a conventional application, but from understanding the interaction characteristics of the new platforms. Ensuring general usability for a particular interaction platform without rigorously specifying envisaged usage contexts helps us to design an artifact that does not restrict the possible application contexts and yet is usable enough to help brainstorm its more exact place for future exploitation. With the themes on Multitouch, I will give examples of design effort currently underway with mobile and Multitouch Wall applications, and also raise (possibly) controversial issues in designing such novel applications when there is no user-base or work practice we can base our knowledge on.