FIX(positional-audio): Fix Source Engine plugin not working on Windows

[davidebeatrici]

The bug was introduced in #4536.

It was only reported last year and because of a side effect: consistent lag for the entire application.
That's something that should be dealt with independently.

[Krzmbrzl]

Lol - that's an easy fix ^^

[Hartmnt]

Is there any reason in particular why you chose to zero initialize the bool in the windows case instead of constexpr bool isWin32 = false? Just curious

[Krzmbrzl]

Right.... An explicit initialization to false might be better.

[davidebeatrici]

Is there any reason in particular why you chose to zero initialize the bool in the windows case instead of constexpr bool isWin32 = false? Just curious

The lack of constexpr is due to the codebase sticking to C++11 at the time.

Right.... An explicit initialization to false might be better.

You mean in the Linux case? I didn't do that to emphasize that we rely on the value that is set later, through HostLinux::isWine().

[Hartmnt]

You mean in the Linux case? I didn't do that to emphasize that we rely on the value that is set later, through HostLinux::isWine().

ah that makes total sense

[Krzmbrzl]

You mean in the Linux case? I didn't do that to emphasize that we rely on the value that is set later, through HostLinux::isWine().

Not sure this really emphasizes something. Generally, uninitialized variables are only a source of bugs - now or in the future. I think for bool we should be save, but explicit initialization shouldn't hurt either way...

[davidebeatrici]

Consider that global variables are always initialized to zero though.

[Krzmbrzl]

Yeah, but the initialization rules in C++ are finicky and thus I think it's better to always be explicit.

[davidebeatrici]

Done.

[Krzmbrzl]

Feel free to merge

[Krzmbrzl]

💚 All backports created successfully

Status	Branch	Result
✅	1.5.x

Questions ?

Please refer to the Backport tool documentation and see the Github Action logs for details