----- Original Message -----
From: Patrice Colet colet.patrice@free.fr To: pd-list pd-list@iem.at Cc: Sent: Friday, June 22, 2012 2:09 PM Subject: Re: [PD] dynamic patching woes...
[...]
Also, I don't really get why you don't use [throw~] [catch~] [send~] and [receive~] instead of [inlet~] [outlet~]
Colet Patrice
That just substitutes one problem for another. The OP wants to encapsulate readanysf~ in an abstraction that provides outlets for the correct number of channels. If one uses nonlocal signal sends then you must deal with Pd's crude scope, either by forcing the user to provide $0 as an arg (ugly), or using an external object to get the parent zero from inside the abstraction (not compatible with Pd vanilla). And now you have the additional problem of a user of a visual programming environment instantiating an object that gives no visual clues as to the data coming out of that object.
(Of course one could just forget $0 altogether and use global send/receive names. But then one better not reuse a send/receive name since none of those objects will work correctly with multiply defined symbols at both ends.)
I'm not saying dynamic patching + [initbang] for variable intlets is a cakewalk, but it's really no worse than the hoops one must go through for other aspects of Pd that are just as ingrained and clunky.
-Jonathan