"...it with" vs "...it to" and "action of [stored action]"

more on stored actions.

This compiles:

lab is a room.

marbles is an animal.
the player is marbles.
box is a container.
gadget is a thing.

a widget is a kind of object.
a widget has an action called the reasonable behavior.
a widget has an action called the unexpected flop.

frotzing it to is an action applying to two things.
frobbing it with is an action applying to two things.

the frob is a widget.
understand "frob" as the frob.

[this is the interesting part]

the reasonable behavior of frob is the action of marbles frotzing the gadget to the box.

[end interesting part]

If I replace the intersting part with this:

[this is the interesting part]

the unexpected flop of the widget is the action of marbles frobbing the gadget with the box.

[end interesting part]

The project no longer compiles, and this failure message prints.

Problem. You wrote ‘the unexpected flop of the widget is the action of marbles frobbing the gadget with the box’ : but that is a very peculiar property value, and ought to be something more definite and explicit.

Is there something I have missed or mistyped here? Or is there something wrong with action names ending with with? It doesn’t seem I can store an unlocking it with action in this way, either.

I think it might be a compiler issue.

As a workaround, you might try using a run-time assignment, as in:

When play begins:
	now the reasonable behavior of marbles is the action of marbles frobbing the gadget with the box.

That definitely works. It’s a shame that I already have so much going on when play begins. I’ve been consciously avoiding putting more things there.

This is a bug, then?

edit:
As a fun bit of trivia: “very peculiar property value” has never before appeared on this forum, so far as I can tell.

Yeah, I can’t see any explanation for this beyond “compiler bug”.

Should I do a report? I don’t mind, but I think I miscategorized the last one I submitted.

I say go for it. If end users can figure out the cause of a bug, that’s great, but if they can’t, the compiler maintainer is the one with the expertise for it! You’ve provided an example that consistently reproduces the bug; that’s already more than most users do.

OK, Done! I’ll mark the workaround as a temporary solution.