[PD] Cyclone in vanilla?

Frank Barknecht fbar at footils.org
Tue Apr 22 20:37:42 CEST 2008


Hallo,
Andy Farnell hat gesagt: // Andy Farnell wrote:

> Unfortunately I've used [pow~] in dozens of other patches and it's 
> quite unfeasible to go back and rewrite all of them and the accompanying
> text. It would take me weeks, and so I feel (on an emotional level)
> quite pissed off because adding [pow~] to vanilla Pd is only a matter
> of will and possibly 10 mins work to push it into the next build.

If you want to avoid too much search-and-replace editing, maybe you
could introduce your own wrapper abstraction version of [pow~] with
[expr~ pow($v2, $v1)] inside? Call it [andypowell~] and do a
search/replace session. Or call it [pow~] and tell people, that they
either use the wrapper or install Cyclone or wait for a
math.h-enhanced Pd-vanilla.

The only thing left to check would be if you ever used [pow~ ARG] with
an argument and maybe make that into a different abstraction. It's not
totally beautiful, but well, at least it's possible to move back and
forth a bit (i.e. if during your writing of the book Miller includes
[pow~] you can just delete the pow~-abstraction paragraph.)

> If I'm going to aim this at Millers Pd rather than Extended then I feel
> it's only fair to have some movement making these small but vital
> improvements to vanilla.
> 
> 
> > Note that I also think, the math objects (abs~, pow~ etc.) should be
> > part of Pd, and probably symbol2list.
> 
> Two of us doesn't make a concensus, but I've got the feeling most would
> agree. 
> 
> Can we make this a catalyst to get a definite commitment to patch
> up vanilla with the missing essentials? I still can't find the 
> message, but I'm sure Miller said something about bringing Cyclone
> into vanilla.

He mentioned on pd-dev, that he has considered this. Quote:

  there's text-editor code in Krzysztof Chaya's library, that I've
  wanted to glom into the vanilla Pd source for some time now (exactly
  so that people can pop up text editor windows for any 'binbuf'
  contents).
  
  Only thing holding me back is two minor issues:  1. I can't decide
  whether it's appropriate to glom te whole of Cyclone into Pd; and 2.
  assuming I don't do that, I'm worried it might break cyclone itself
  to export symbols from Pd that are also defined by cyclone.

  http://lists.puredata.info/pipermail/pd-list/2008-03/060677.html

> Can I please ask all the maths heads here to help define what would
> constitute a mathematically complete object set for audio signal processing?
> 
> Right now my 'missing' list includes [z~], [abs~], [ln~], [log~],
> [pow~], [tanh~], [cosh~]

I think, most operations from the standard C math library should be
included as (signal) objects in Pd. The selection Lua took would be my
guideline: http://www.lua.org/manual/5.1/manual.html#5.6

Note that there's no [ln~] here or rather: the natural log. would be
called [log~] if following math.h. And as known [cos~] doesn't follow
math.h nor [cos] either.

Ciao
-- 
 Frank Barknecht                                     _ ______footils.org__




More information about the Pd-list mailing list