On Fri, Feb 03, 2012 at 07:54:07AM +0100, David Henningsson wrote: > I think your solution with two different profiles is correct > (assuming your analysis of the problem is correct). > Having input paths affecting the output side and vice versa is very > uncommon in the desktop world. Had it been common, maybe things > would have been differently modelled. It's pretty uncommon in the embedded world too, you'll often see the input and outputs all being mutually exclusive (or only able to take the same signal) but it's very unusual that the output would force a particular input. It is common for systems to do this from a use case management point of view (always using the headset if it's present for example) but that's not normally a hardware restriction.
