mlx_destroy_display leaves memory unfreed

[UnWaDo]

mlx_destroy_display was added to free memory, allocated with mlx_init. Unfortunately, now it leaves t_xvar *xvar unfreed, so memory is still leaking

[ggjulio]

I kinda agree.
mlx_destroy_display should never have existed/merged because it's not the proper destructor of mlx_init.

Then students break the Soc principle by freeing themselves the object returned by mlx_init. In order to "have 0 leaks".
I know most of studs thinks still reachables reported by valgrind are bad things that must disappear at all cost.
But must of the time it is okay to have some and don't deserve a leak flag.

For example as OL said, (author of mlx)
the still reachable "leak" of mlx_init doesn't matter since all our program's lifetime ends when graphics ends.

However a proper destructor would have been necessary if our program continue after graphics ends.
(to release the memory right when we don't need it anymore)

Also mlx_destroy_display is a complicated name for what it does.
Probably mlx_quit would be better.
Like the SDL -> SDL_Init() and SQL_Quit()