a way to create a big-bang3d world as a child of an existing frame (or other gui obj)?

[AlexKnauth]

Is there a way to create a big-bang3d world as a child of an existing frame (or other gui obj), and if not, could an option for that be added?

I'm thinking about this because I wanted to add a scoring system to my-unpossible where it would display the current score in the same window as the big-band3d world.

[ntoronto]

That's an interesting way to work around two current limitations: Pict3D doesn't have a way to do 2D yet, nor can it render in layers. (It will eventually.)

Rather than adding a parameter to big-bang3d, I think I'd rather provide the canvas that big bang programs run on, and let users use it to control how programs run and render.

[AlexKnauth]

But would that allow a big-bang-like interface?

[ntoronto]

I think so. A big-bang3d-like function could return a canvas, which could have start and stop methods that receive and return the game state, and possibly methods for changing parameters like #:frame-delay.

As far as remaining purely functional goes, that kind of gets thrown out the window as soon as you do GUIs in Racket, so I'm not worried about it.

[AlexKnauth]

What do you mean? Do you mean that you would make another big-bang3d-like function that takes a parent gui object in addition to the normal #:on-draw etc, and returns a canvas immediately, which would then require a call to a start method? Or were you thinking of something else?

But for purely functional, the handlers for updating the world-state and drawing the scene would still be functional, right? And the start method would be imperative? But does it even need a start method, or could it use something like on-superwindow-show to do that? And without that, does it even need to return a canvas in the first place?

Unless you want to provide imperative methods for other reasons, returning a canvas shouldn't be necessary, or am I wrong? So then the new big-bang3d-like function would behave exactly like the current big-bang3d function except with this extra argument, so then you might as well make it an optional argument to the current big-bang3d function?