End the funking song I'm going insane

[Lasercar]

Does this PR close any issues? If so, link them below.

Fixes #4304, (finally) fixes #3142

Briefly describe the issue(s) fixed.

#4309 (comment)

@Hundrec

Has found the way to replicate the bug - when the song ends, click off the window, then click back into it - what ends up happening is that you bypass music's oncomplete function (because the music still plays even when you've clicked off) and so, the song doesn't end.

This also means that once the music playing when unfocused haxe issue is fixed, it should fix this issue too (but then again, maybe keep this in just in case it's also caused by something else?).

Include any relevant screenshots or videos.

[Lasercar]

[UH-OH!] yeah!

[KoloInDaCrib]

Your version does not fix the original issue.

2025-03-16.08-16-25.mp4

For your information, I am not running anything else besides your pr, otherwise there would be visible changes.

I've even limited myself to using your preferred FPS cap of 80 without any offsets.

I honestly have a feeling you are purposefully sabotaging my pr to get yours merged and mine canned.

[JackXson-Real]

Your version does not fix the original issue.

2025-03-16.08-16-25.mp4

For your information, I am not running anything else besides your pr, otherwise there would be visible changes.

I've even limited myself to using your preferred FPS cap of 80 without any offsets.

I honestly have a feeling you are purposefully sabotaging my pr to get yours merged and mine canned.

He’s just trying to help. Apparently your PR hasn’t fixed it either, so at this point a fix for this bug is fair game. We all want this bug to be fixed, don’t assume someone is doing something to sabatoge you, especially on GitHub, this is not a competition.

[KoloInDaCrib]

He’s just trying to help. Apparently your PR hasn’t fixed it either, so at this point a fix for this bug is fair game. We all want this bug to be fixed, don’t assume someone is doing something to sabatoge you, especially on GitHub, this is not a competition.

I don't have a problem with him "taking my spot" or whatever, I mostly have a problem with him claiming that my pr doesn't fix the issue, even though I have proven it does. Besides, you only speak for his name, but have you tried both prs and judged them based or not they fix the original issue or not?

[Lasercar]

Your version does not fix the original issue.
2025-03-16.08-16-25.mp4

For your information, I am not running anything else besides your pr, otherwise there would be visible changes.

I've even limited myself to using your preferred FPS cap of 80 without any offsets.

I honestly have a feeling you are purposefully sabotaging my pr to get yours merged and mine canned.

Look, I did say in your PR that

If this doesn't end up fixing it, then maybe I'll have a go at fixing it too.

So I did? Was that the wrong thing to do?

And besides, because the bug can be reproduced on literally every fix everyone's made, it's basically fair game at this point.

IDK.

[Lasercar]

Oooh, I got again with my changes!

And my breakpoint on end song at if (event.eventCanceled) return; didn't fire, so something very funky is going on.

Ah, so the songPosition is being capped at the music music length, so all I need to do is add an equals (I hope).

[trayfellow]

Funkin' github ranked

[Hundrec]

Changes here integrated into #4309

[Lasercar]

Superseded by #4334

(if you don't mind or can fix the weird issue hundrec found)

[KoloInDaCrib]

Noticed a bug with this while playtesting - if you have a positive offset, it will cause a null object reference to appear in the results screen (in the video I used visual offsets of 100ms)

2025-03-16.16-14-42.mp4

I assume this is because you use Conductor's songPosition variable (which has offsets accounted for)

which makes it so that endSong is called twice, and therefore the game moves to results screen twice (proven here)

In comparison my PR uses music's raw time (no offsets) and it the game works without any NORs

2025-03-16.16-03-50.mp4

In the future I recommend not using breakpoints, since by that you're limiting yourself to seeing what should happen instead of letting the game run to see if your changes have any unintended consequences

[Hundrec]

See Kolo's comment above

[Lasercar]

Noticed a bug with this while playtesting - if you have a positive offset, it will cause a null object reference to appear in the results screen (in the video I used visual offsets of 100ms)
2025-03-16.16-14-42.mp4

I assume this is because you use Conductor's songPosition variable (which has offsets accounted for)

In comparison my PR uses music's raw time (no offsets) and it the game works without any NORs
2025-03-16.16-03-50.mp4

In the future I recommend not using breakpoints, since by that you're limiting yourself to seeing what should happen instead of letting the game run to see if your changes have any unintended consequences

Fixed it!

That was easy (and before you ask, no, I did not make the same changes as you):

2025-03-17.15-42-42.mp4

[EliteMasterEric]

Gonna resolve the merge conflict with #4309 and merge this too.