Initial touch event support #60
Conversation
include/simple2d.h
Outdated
| @@ -130,13 +135,16 @@ typedef struct { | |||
| int direction; | |||
| int axis; | |||
| int value; | |||
| int64_t finger_id; | |||
| int64_t touch_id; | |||
There was a problem hiding this comment.
Instead of int64_t, do you think these should be of type SDL_TouchID and SDL_FingerID, respectively?
There was a problem hiding this comment.
Its not done for SDL_JoystickID either, but I don't think its a big deal except it leaks SDL into Simple2D more than is already done
There was a problem hiding this comment.
Huh, didn't even realize joystick ID was an SDL type. Guess I must have copied some code somewhere using int. Looks like they use SDL types for future proofing, but I think it makes sense as a principle not to expose SDL types to users (the ones we use in S2D types are all for internal use) so using regular types make sense here. Might actually change to long long if you're ok with it.
src/window.c
Outdated
| } | ||
|
|
||
| }break; | ||
| case SDL_FINGERDOWN: |
There was a problem hiding this comment.
Since SDL_FINGERDOWN and SDL_FINGERUP are identical, save for event.type, I might suggest combining both events like in mouse down/up here. (Reminds me, we can probably do the same for SDL_JOYBUTTONDOWN and SDL_JOYBUTTONUP as well.)
src/window.c
Outdated
| fy = e.tfinger.y * window->height; | ||
| dx = e.tfinger.dx * window->width; | ||
| dy = e.tfinger.dy * window->height; | ||
| S2D_GetMouseOnViewport(window, (int)fx, (int)fy, &mx, &my); |
There was a problem hiding this comment.
I'm a bit confused by the multiplication of the window dimensions. Does S2D_GetMouseOnViewport not correctly translate the points in the window? (Since we're using it for more than mouse now, should probably rename it to S2D_GetPointOnViewport or something.)
There was a problem hiding this comment.
Touch events seem to deliver not positions, but indications of how far in a given dimension the event occurred, normalized between 0 and 1
e.g. 0.5 of x, 0.3 of y
which then need to be translated into coordinates in the window/viewport
The TODO comment "Viewport in scaling" seems to be not relevant - it seems a consequence of scaling the window to a bigger width is then that touch events can occur at a -ve position - since the original 0 axis is no longer flush with the left of the display, that is therefore OK and I had simply gotten confused.
There was a problem hiding this comment.
Ah, didn't realize that, thanks. Running the test card, my MacBook actually triggers touch events as well, I guess because the touchpad is technically "multi-touch." The x and y coords are off, so I'm going to investigate why that is.
|
any updates? Will it be merged soon? |
|
any updates on this? |
|
Another bump for merge |
Pull request for the patch in #59