[PD] Re: PD-list digest, Vol 1 #1065 - 14 msgs
cgc at humboldtblvd.com
Tue Mar 30 20:21:59 CEST 2004
On Mar 30, 2004, at 8:36 AM, Stefan Turner wrote:
> Any chance of some more info on how this might work?
> Would there be a new object to apply any
> transformation to vertex positions and colours, or
> would it be based on getting and setting arrays? I
> think both could be quite useful, with the
> 'transform-vertex' (or whatever) object being
> (presumably) more efficient and ok for simpler stuff,
> and arrays for anything more complicated.
Nothing about vertex arrays is set at this point. Right now it's just
a collection of objects on my HD that I'm trying to make into a
coherent set of tools for 'dynamic geometry'. The current idea is to
have a set of vertex generators like a grid or a .obj model loader and
a set of vertex manipulators for color, position and so on. The idea
is to process vertex, color, normal and texture coordinate data just
like pixels or audio. Pd/GEM is not a very effective environment for
representing large data sets to the user, so very fine editing of the
data might not be all that possible. The goal is to have a way to
algorithmically alter geometry over time - think about something like
newwave or Cyrille's pmpd applied any model you load into GEM and you
start to get the idea. There will probably be lots and lots of of
problems to solve before these objects become all that useful.
> Sorry for all the questions, but this sounds just like
> what GEM is currently lacking (for my needs). Sorry if
> any of this has been answered: I haven't been able to
> find much about it.
By all means let us know what you find lacking about GEM. The past
year has been spent getting 'the basics' on each platform up to par and
improving performance and stability. Now it's time to get some really
interesting things going.
More information about the Pd-list