[PD] "declare" object : "wrong" -path retrieval for embedded abstractions ?

pierre cage pierre.cage at gmail.com
Thu Nov 9 23:55:29 CET 2006


Hi,


As an exemple, here are three abstractions and a directory tree :


----------------------------------------------
./ abs_0.pd
----------------------------------------------
#N canvas 87 210 474 324 10;
#X declare -path abs_0;
#X obj 94 107 abs_1;
#X obj 94 76 declare -path abs_0;
----------------------------------------------
./ abs_0 / abs_1.pd
----------------------------------------------
#N canvas 46 88 474 324 10;
#X declare -path abs_1;
#X obj 93 107 abs_2;
#X obj 94 76 declare -path abs_1;
----------------------------------------------
./ abs_0 / abs_1 / abs_2.pd
----------------------------------------------
#N canvas 120 180 470 320 10;
#X text 59 26 dummy abstraction;
----------------------------------------------


When opening abs_1.pd, abs_2.pd is found. But if opening abs_0.pd,
abs_2.pd isn't found (as searched for in ./abs_1 instead of
./abs_0/abs_1).


Cheers ... and let me give Günter his due,


Pierre Cage


2006/11/9, Hans-Christoph Steiner <hans at eds.org>:
>
> On Nov 5, 2006, at 2:35 PM, pierre cage wrote:
>
> > Hi list,
> >
> > I was stupidly thinking of a standard (but redundant) dir hierarchy to
> > store dependent abstractions : NAME.pd should contain ( declare -path
> > NAME_abs ) and /NAME_abs/ should host embedded abstractions,
> > themselves containing ( declare -path ... , etc.
> >
> > Unfortunately, this can't be done as umbedded "declare" objects miss
> > relative dirs to point unanimously to the main dir (i.e to the path of
> > the "patch").
>
> Hmm, that sounds like a bug, if I understand it correctly.  Could you
> send examples?
>
> > Comparatively, HCS's "getdir" object does a good job retreiving each
> > relative path.
>
> Just to set the record straight, the illustrious Günter Geiger wrote
> [getdir], not me.  I am just an avid user and proponent of it.
>
> .hc
>
>
>
> ------------------------------------------------------------------------
>
> All information should be free.  - the hacker ethic
>
>
>
>
>




More information about the Pd-list mailing list