Stash time in the fluid state?

[MTCam]

This is a feature request suggestion, mostly to see if it is laughed out of the room.

Can we stash time somewhere in the fluid state? That'd be great? 🤔

Originally posted by @MTCam in #648 (comment)

[majosm]

One alternative would be to define the BCs inside the RHS. Then you wouldn't need to pass the time through the operators, you could just evaluate the boundary solutions at the current time and set up the BCs accordingly before passing them in.

[inducer]

@majosm beat me to suggesting that. :)

[MTCam]

I think I'm unclear on the details of how that would work. Wouldn't that get a a re-compile at every substep?

[inducer]

No, I don't think so. Time comes into the RHS as a placeholder for an argument. You partial that into the relevant BC functions and pass those into the operators. Nothing should trigger a recompile there IMO.