Application does not detect mouse motion events on macOS

[asandroq]

Problem

Right after startup the application does not respond to mouse motion events. Clicking to start the ball works but the ball goes straight up and leaves the window through the top. All keyboard events go to the terminal that started haskanoid.

System info

macOS 10.13.15
Darwin lorien.local 17.6.0 Darwin Kernel Version 17.6.0: Tue May 8 15:22:16 PDT 2018; root:xnu-4570.61.1~1/RELEASE_X86_64 x86_64

GHC: x86_64-osx-ghc-8.0.2
Clang: Apple LLVM version 9.1.0 (clang-902.0.39.2)
SDL: 1.2.15

[ivanperez-keera]

I don't see anything in the code that would hint to why this may be happening.

Are you getting mouse motion events? (can you modify Input.hs to print them if they occur).

Can you somehow add a print here:

https://github.com/ivanperez-keera/haskanoid/blob/master/src/Input.hs#L232

So that, if you are getting mouse motion as a different event, you can see what it is?

Does this program work for you?

https://github.com/keera-studios/haskell-game-programming/blob/zuriHac2018/tutorials/sdl/sdl1/input/example1.hs

[asandroq]

There no other different events being fired except for LostFocus and GetFocus events. After some more investigation, there seems to be two issues here:

1. The MouseMotion SDL event seems to be only detected if the trackpad is pressed.

2. The calculation of the velocity in Game.hs:519 is returning invalid values for some reason, always very small floating point numbers close to 0.0.

(-2.5236576396411892e69,0.0)
(6.309144099102973e69,0.0)
(-1.388011701802654e70,0.0)
(2.914824573785573e70,0.0)
(-6.704096519706819e70,0.0)
(1.4749012343355003e71,0.0)
(-3.5397629624052006e71,0.0)
(6.72554962856988e71,0.0)
(-1.2778544294282774e72,0.0)
(2.427923415913727e72,0.0)
(-5.341431515010199e72,0.0)
(1.2819435636024478e73,0.0)
(-2.8202758399253852e73,0.0)

[asandroq]

When trying the input example I still see problem 1, so it seems to be an SDL problem. But after clicking, I can move the ball without problems.

[chriz-zeller]

This issue contains two issues. We split them in #70 and #71, and close this issue.
Thank you @asandroq for mentioning and helping to debug them.