...but I would very much like to close it with a bang as well - the bang would output first to cue up the pattern I've entered and then close the window as well.

DO NOT go querying on this, you won't find anything.

I'm open minded about a shell command if someone seconds my opinion that it may be the only way. I would prefer an internal message or object though.

Thanks,
J.P.

[;pd-test.pd menuclose(

mvh

It worked smoothly the first time, prompted me to save changes, great.

Subsequent trials, closing after I have not made any changes, exit PD altogether.

Has anyone else tried this? It's really too bad that there is so little documentation on internal messages.

You can't do it from the patch you are closing. You have to do it from another patch.

And if you try to close two patches with the same name pd will crash.

Another thing: if you want to suppress the save changes message put a 1 after menuclose like [pd-blabla.pd menuclose 1(

Lack of documentation is partly because this use of internal messages is not supported.

Anyway attached is a small example. Open main.pd and start the metro

mvh

http://www.pdpatchrepo.info/hurleur/main.pd

You can't do it from the patch you are closing. You have to do it from another patch.

And if you try to close two patches with the same name pd will crash.

Another thing: if you want to suppress the save changes message put a 1 after menuclose like [pd-blabla.pd menuclose 1(

Lack of documentation is partly because this use of internal messages is not supported.

Anyway attached is a small example. Open main.pd and start the metro

mvh

http://www.pdpatchrepo.info/hurleur/main.pd

I had to make sure that the windowcloser.pd was not a subpatch of my main patch, of which the popup window was also a subpatch.

I still wanted to send from the same bang that applied changes in my popup window. It crashed PD, even though the windowcloser.pd patch, where the [r closeit]----[pd-popup.pd menuclose( command fires, is not a subpatch of the main patch.

So, I added a delay in the windowcloser.pd patch between it receiving the bang and actually hitting the message-command box (50 ms, excessive but who cares); this seems to allow me to close my popup window with one click in the very window that I am closing. Yea!

Of course windowcloser.pd must run in the background this entire time, but I imagine it will be able to provide similar functionality for any number of my subpatches...now how much I/O glitching is this opening and closing of windows going to cause me? I don't want to know.

• J.P.