Fixed issue where mouse clicks were always reported as separate button down and button up events.

[Lyle-Tafoya]

SDL_PollEvent was always returning 0, which meant we never set action = BUTTON_CLICKED. Using SDL_WaitEventTimeout instead of napms followed by SDL_PollEvent seems to fix this.
I mentioned this bug in issue #117.

[Bill-Gray]

As noted under issue #117, I'm not seeing the problem. I am reasonably sure that not getting mouse clicks would be breaking various bits of code and would be reported... but there are so many variations between SDL 2.x.x and SDL 2.y.y that I could easily imagine the problem you report showing up in a version I'm not using.

The fix you propose is quite straightforward and works on my machine. I don't see any reason it wouldn't work on other machines, so I'm pulling it.

That said... I am wondering if the underlying problem is that napms() is broken on the SDL2 you're using? If you call napms( 2000), for example, do things pause for two seconds? (If not, we've yet another bit to fix...)

Thanks!

[Lyle-Tafoya]

Yes, if I call napms(2000), it does indeed pause for 2 seconds (and the click is still reported as separate button press and button release events). I had reached the conclusion that sleeping must interfere with the ability for input events to be processed, which makes me more confused now that I know you were unable to reproduce this problem. I am using the version of SDL that PDCursesMod builds automatically (2.26.2)