@jameslo I would use [coords( rather than [donecanvasdialog( as it is less "dirty". [pdcontrol] I have never tried.
I can't remember whether you need to exit from edit mode after the [coords( message.
David.

@jameslo Yes, I think so too......
It seems to have a tilt to archaeology... digging up blasts from the past.
I will have to put my "bad cop" hat on again.
David.

@nicnut Yes, as the [phasor~] output jumps it could output samples that jump from near 0 to near 1 or vice versa and that will produce clicks.
You could window each grain, or as I did here soft duck the output during the jump.
The effect is the same I think.
Microsound-granular.pd
See [s smooth] and [r smooth] for the duck.
David.

@crttrkix P.S. As well as passing arguments from [cue] to the name of the abstraction it contains, you can also pass arguments to it.
i.e. [cue 1 22 33 44]
contains
[whatever$1 $2 $3 $4]
where $2 $3 $4 inside [whatever1] resolve to 22 33 44 upon creation.
David.

@crttrkix "This demonstrates a bit of an inconsistency in their behavior (I don't know why this is) that is good to know.

It is just that [whatever] is an abstraction... that is another patch..... whatever.pd.... inserted into your patch.
[pd whatever] is a window within the patch that contains it..... useful for keeping a patch tidy but also to fix an order of operations and to set a different block size for resampling, and for processing audio at a different block size (right down to single sample processing).

And what @jameslo said..
David.

@crttrkix You can pass arguments, but not as part of the name.
$0 is not so useful, as when you want to send messages from outside the abstraction you can get it to declare its $0 value but that becomes complicated.
So it is easier to give each instance known arguments.
Arguments are passed down to sub-patches within the abstraction.
So probably.....

I have used the first argument directly in [expr]
Since some version of Pd the $1 would be automatically escaped within the formula of [expr] and made literal, so I have escaped it again to make it a variable as required in this case.

You can have as many arguments as you wish..... intteger and string....
Inside the abstraction use [symbol $1] to loadbang a string.
I tried to make a tutorial here.... https://forum.pdpatchrepo.info/topic/9774/pure-data-noob/4

For the sub-patch it is not actually useful as the sub-patch is just a window within [cue].... so all you have done is change the name of the sub-patch.
If you want a completely different sub-patch in each instance of [cue] then use another abstraction (a patch) instead.
[cue 1 34]

inside [cue] $1 becomes 1 (in an object) and [patch$2] becomes [patch34] which you have created as patch34.pd

But if the sub-patches are identical then pass arguments from the creation of [cue]..... as many as you need...
[cue 1 34 woof lala 12 rev delay]
Inside object names they will be resolved...... and not escaped.
David.

Should be here..... https://puredata.info/docs/developer/PdFileFormat ...... but the whole site is unreachable right now.
It used to be included in Pd extended source code..... another loss....
Here it is....... pdfileformat.zip
David.

@pepika32 [gemwin] should be an object..... not a [gemwin( message.
Hopefully it will then work as you expect.
David.

@Moothart The easy way will be to use [print] and you will see the value in the console even though a number box might not have time to show it.
Or you could use [sel 2 5 6 7 10] with bangs on the first 5 outlets and see the effect in the patch.
David.

@Moothart If timing is a problem..... and it is always a double bang..... then you can solve it like this using a [toggle]....
David.