I had an idea and want to ask about the opinion of the developers.
In general the current implementation of desmume has only a mouse input for the touchscreen.
However many games use the touchscreen as button replacement. The idea is to map the touchscreen buttons via coordinates to the keyboard or to a controller.
For example New Super Mario Bros
To use the mushroom the player has to press the mushroom button on the touschreen. However after a while it can be annoying to use the mouse or a pointing device to select this button.
I tried with GlovePie to simulate several fake mouses to click on fix coordinates with the buttons on a controller. It was working pretty well. However if I move the window of desmume or if I switch to fullscreen the solution from GlovePie is gone. Furthermore GlovePie is a dead app and the development ended years ago. It was more or less only a feasibility.
But the idea is to use the same approach from GlovePie, however integrated in desmume. Means, the user defines a button which shall be used as a click event on touchscreen coordinates. Instead of using a mouse to click on the touchscreen, the user can press a button on the controller (for instance the dpad on the Xbox 360 controller) to click on pre-defined touchscreen coordinates. The resolution of the screen (256 × 192) and the coordination system are known. Hence, the benefit would be that the defined coordinates stay the same, if the window from desmume is for example moved. So the mapping of a button to a touchscreen coordinate, would always lead to the same result.
Especially games like New Super Mario Bros, Super Pricess Peach and other games would become more playable.
However I am not so familier with the code base of desmume and the efforts to implement the idea.
Last edited by papermanzero (2016-05-26 14:23:44)
Youre not the first person to have thought of this. You can automate with authotkey.
I know I could use Autohotkey. I could also use AutoIT.
And if I am very ambitious I could also C and C++.
But this is not my question. I see more benefits to integrate it directly in the emulator instead of an external application.
I already saw some thread about past proposals which are the same.
It seems you implemented that feature already for MacOSX.
I am really lazy to search through sourceforge (I am more a fan of github), but I am interested about the obstacle concerning a windows port of the feature.