Many thanks to you all,
[initbang] seems just what I was looking for. As for individual, dynamic firing with [loadbang], the problem with

[loadbang(
|
[send pd-foo.pd]

is that I have several foo abstractions and I would like to fire them individually. So, by now [initbang] seems the way to go.
Josep M


> From: reduzent@gmail.com
> To: pd-list@iem.at
> Date: Sun, 24 Mar 2013 12:25:35 +0100
> Subject: Re: [PD] some issues with dynamic patching
>
> On Sam, 2013-03-23 at 15:39 +0200, Alexandros Drymonitis wrote:
> > Concerning [loadbang] you should use [initbang] instead AFAIK. But
> > that's not vanilla.
>
> There are two separate issues to be considered:
>
> [initbang] should be used when you dynamically create xlets within an
> abstraction, so that those are created before the connections of the
> parent are drawn.
>
> I think what OP means is that [loadbang]s in dynamically created
> abstractions do not fire too late, but not at all. Whether this is a
> feature or a bug is not clear to me, but it is the current behavior
> which has been discussed many times on this list. To me this behavior
> actually makes sense. It allows you to first create many instances of
> the abstraction dynamically and only then let them loadbang by sending a
> 'loadbang' message to their canvasses.
>
> Roman
>
>
>
>
> _______________________________________________
> Pd-list@iem.at mailing list
> UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list