group id exists but yet not . . .

[Dr. Hawkins]

On Sun, Jan 4, 2015 at 4:14 PM, J. Landman Gay <jacque at hyperactivesw.com>
wrote:

> It's been that way since I can remember. If the group is placed on the
> current card, you can refer to it as "group", otherwise you need to use
> "background". This applies even to closed stacks, so maybe when it worked
> for you before, the current card of the stack happened to have the group on
> it.
>

That can't possibly be it:  I stumbled across this in a routine that loops
through half a dozen substacks, every card of each, and every control,
referencing every one.  When run, it is typically only the main stack that
is open.

And shouldn't that apply just as much to

group a of group b of card c of stack d

which worked, as

group a of stack d

which didn't work.


> This is an invalid reference:
>
> put exists(background id 1004 of background id 1111 of card id 1016 of
> stack "nv")
>
> Since every group has a unique ID, you don't need to specify the enclosing
> group, only the specific background:
>
> put exists(bg id 1004)
>

What's confusing me here is not whether or not it *should* work (which is
its own set of issues), but that there is a difference between these two.
They were:

put exists(group id 1004 of group id 1111 of card id 1016 of stack "nv") &&
   exists(group id 1004  of stack "nv")

both refer to a group on the same card, but one works, and not the other.

-- 
Dr. Richard E. Hawkins, Esq.
(702) 508-8462