I think doing more than one UI upfront is an optimistic mistake. Pick a primary platform and optimise it. Maybe do a less-than-awesome port, if you're eager. I

I think doing more than one UI upfront is an optimistic mistake. Pick a primary platform and optimise it. Maybe do a less-than-awesome port, if you're eager. If, down the road, your experience with the primary platform leads you to believe there is a large market for the taking, then you can weigh up improved sales versus development effort for a truly excellent implementation on another platform. Regarding Qt on the Mac: I've done it before. It would be my less-than-awesome port. If you're not using Cocoa on the Mac, making a UI on par with a "native" app is a sisyphean task: you're fighting uphill and you never quite get there.