Here are some revisions.
cnv-adr-testmain-rev-normal.pd: Here I avoid dynamic patching altogether, because it's not necessary for the behavior your trying to get in the abstraction.
cnv-adr-testmain-rev-dyn.pd: Here I added [namecanvas $0-cnv] and sent to that instead of the abstraction name. You
still get the cord from inside the abstraction showing up on the parent. Looks like a bug.
Sending to pd-abstractionName.pd will send to all instances of an abstraction (desirable in some circumstances).
Using [namecanvas $0-whatever] will bind a unique symbol to that canvas, so you can use it to send only to "this"
instance using [s $0-whatever]. (The [iemguts/sendcanvas] object does the same thing, with the added benefit that
you don't have to actually bind a symbolic name to the canvas.)
-Jonathan
----- Original Message -----
From: Michael Zacherl. sdiy-mz01@blauwurf.info To: PD list pd-list@iem.at Cc: Sent: Wednesday, August 31, 2011 10:33 AM Subject: [PD] multiple instances of the same dynamic abstraction?
Hi, I'm currently experimenting with dynamic patching, sending mgs to cnvs etc.
One obvious question is, how to address a dynamic abstraction when having multiple instances of it in a main patch. Is there a way around this? Besides that any msg to the cnv hits all occurrences of that particular abstraction I manage to crash pd almost regularly with 0.42.6, ext-0.42.5 and 0.43.0. 0.43.0 also shows an Tcl-error message, probably regarding the line which sticks out when running that example on 0.42.6 and ext-0.42.5. (this cord also shows up when using just one instance of this abstraction).
I included a simple example. Any hints?
Thanks a lot, Michael.
-- keep your ears open: http://blauwurf.at http://soundcloud.com/noiseconformist
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list