So I’ve been thinking about this as Matt has brought it up. Pygame.display.quit turns the screen SDL_surface into NULL— but should it? Is there any (current) reason the screen surface can’t continue in software after the display behind it has quit? Why try and error on everything a surface could do, what if we could instead error on pygame.event.get or pygame.display.flip.

You can probably include stuff from #3525 in your PR. There is also this comment that may be relevant #3525 (comment)

So I’ve been thinking about this as Matt has brought it up. Pygame.display.quit turns the screen SDL_surface into NULL— but should it? Is there any (current) reason the screen surface can’t continue in software after the display behind it has quit? Why try and error on everything a surface could do, what if we could instead error on pygame.event.get or pygame.display.flip.

This is fair, but I still think that my PR is still relevant. If someone somehow still would end up with a weird, broken surface it would still potentially segfault

You can probably include stuff from #3525 in your PR. There is also this comment that may be relevant #3525 (comment)

Are you suggesting that the two PR's should be effectively merged into one? Since the same code is repeated in different places in both, I think that would be a good idea. I also like the idea of a macro. Maybe SURF_INIT_CHECK instead of VIDEO_INIT_CHECK like @Starbuck5 suggested in your PR?

Maybe something like this?

#define SURF_INIT_CHECK(SDL_SurfObj) if (!SDL_SurfObj) {return RAISE(pgExc_SDL_Error, "Surface is not initialized");}