apple2e: fix IIc mouse regression

[arekkusu42]

This PR fixes #14463, which is a regression introduced by 92aa47f.

The "unconnected joysticks now read properly" portion of that commit fixed joystick 1 as advertised, however it incorrectly propagated similar logic to the IIc-specific mouse handling for C066/67. The result is that the mouse stops functioning if no joystick is plugged in.

For folks affected by this in 0.282, a workaround is to launch with -gameio joy, which will restore mouse functionality.

[colinleroy]

Confirming both the regression and fix :)

[cuavas]

For folks affected by this in 0.282, a workaround is to launch with -gameio joy, which will restore mouse functionality.

Why is it possible to have a mouse and joystick connected at the same time at all? Don’t they use the same physical connector on the IIc?

[colinleroy]

For folks affected by this in 0.282, a workaround is to launch with -gameio joy, which will restore mouse functionality.

Why is it possible to have a mouse and joystick connected at the same time at all? Don’t they use the same physical connector on the IIc?

It's not physically possible, indeed: they do share the same connector. But MAME doesn't deactivate the mouse handling when a joystick is connected so far.

[arekkusu42]

They do use the same connector, which is why I had added this code in the first place-- in order to use the mouse, it should be plugged in, no? However, MAME isn't emulating the mouse via a gameio device, instead apple2e_state::update_iic_mouse() is hooked up via analog optional_ioports.

So, there is probably a (preexisting) bug here where if -gameio joy is used, the mouse should not work. I don't have a //c here to test the C066/67 interactions; I only restored the old behavior.