2007/12/30, Frank Barknecht < fbar@footils.org>:

Hallo,
raul diaz hat gesagt: // raul diaz wrote:

> I'm working on a patch as interface for Pulsar Grain Synthesis based on a
> variation of nullpointer abstractions (good work Tom!):
>
> - http://www.nullpointer.co.uk/-/pd.htm
> - http://www.nullpointer.co.uk/-/files/pulse-grain-generator.zip
>
> I'm making some functions (constant, ramp up/down, triangle, sine, gaussian,
> etc) to store on tables for parameters control.
> I've made an abstraction to store an array with selected name which stored
> this functions [p-table1].
> [p-tables] abstraction dynamically create six of this abstractions.

Do you really need to do this dynamically? I don't see any compelling
reason for dynamic patching here.

The [p-tables] abstraction will always be the same patch after it is
loaded. Additionally if you really want to create two or more
instances of [p-tables], like [p-tables 1] and [p-tables 2] your
current approach may break anyway unless you use [namecanvas] and as
soon as you introduce that, I would rather recommend to use a [pd
$0-something] subpatch for dynamic patching, and with that, you could
could then send a message "loadbang" to [s pd-$0-something] to
initiate another loadbang for the newly created objects without
getting into a loadbang-loop, as you would if you use [namcanvas] in
your current patch.

So basically my recommendation would be to forgo dynamic patching in
your case, it does more harm than good here.

(Here I also get lots of crashes when trying to open a created
[p-tables] instances. Anyone else gets this?)

Ciao
--
Frank Barknecht _ ______footils.org__

_______________________________________________
PD-list@iem.at mailing list
UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list