[PD] Cyclone in vanilla?

marius schebella marius.schebella at gmail.com
Tue Apr 22 20:32:22 CEST 2008


Andy Farnell wrote:
> On Tue, 22 Apr 2008 09:52:17 +0200
> Frank Barknecht <fbar at footils.org> wrote:
> 
>> I'm reluctant to mention [counter] here, which also was coded many
>> times, unfortunatly in incompatible ways. I'm reluctant, because
>> [counter] is too basic to be included. 
> 
> 
> I heartily agree. In fact I don't suggest any message domain changes to
> vanilla right now, just the missing signal processing operators. [counter]
> is a perfect example of something that should never be in core Pd, it is
> ill defined. Things like [tanh~] however are fundamentally defined, so there's
> no change of two conflicting implementations (unless one is wrong).

If you look at "counter" as a class of several methods, then you will 
agree, that it should have more features than just adding "1" everytime 
you send it a bang message.
It should be able to count in different steps, start and end at 
different values, should be resettable, and probably should incorporate 
some more features like report when it hits the maximum or when it jumps 
back to the lowest value.

there needs to be a place where the functionality of such kind of higher 
level object classes need to be discussed and specified, and then 
provided as whatever (c-internal, c-external, abstraction-internal, 
abstraction-external). I don't care.

but from the practical viewpoint this is inevitable. everybody that 
codes in pd is dependent on a set of such objects.

please, please, I apply to all pd developers and users to agree on a set 
of higher level objects or "standard" externals. even if they don't get 
shipped with every distribution (although I think they should be shipped 
and incorporated into pd vanilla), everybody should be aware of a canon 
of objectclasses.
this would save hours, if not days of patching time.

putting cyclone in vanilla would be a good start.

marius.




More information about the Pd-list mailing list