Graphics Reference
In-Depth Information
situation at hand. Similarly, the successful evolution of design tools will depend
on whether we can create software that can rapidly and accurately evolve new
geometric representations in response to the design problem at hand.
Some human languages have words to describe emotional states or par-
ticular situations that other languages don't. The complexity and available
vocabulary of a human language inluence what can be said. Similarly, the way
a particular design software tool handles internal geometric representation
determines how well it can manipulate a design.
Most design software today speaks a simple form of language and thinks
like a traditional paper blueprint. It speaks a simple crude dialect of simple
words. For example, design software has two words to describe the presence
or absence of material in a design: “there” or “not there.” There's no capacity
for ambiguity. There's no way to describe a physical object's growth, or change
or conditional adaptation. Like a non-native speaker or a child learning his
irst words, design software simply and baldy states the present details of a
design's shape and materials.
Until design tools can speak more eloquently we can't fully tap into the
potential of 3D printing. There are simple ways for a computer to “think shape”
and there are more sophisticated ways. If I had to rank design paradigms from
simplest to most luid and adaptive, here's how I would rank them.
The simplest geometric representations would be traditional paper blue-
prints, surface mesh models or solid geometries. These design notations
describe a simple, ixed shape. They are the equivalent of speaking a few
simple descriptive words.
Next, somewhat more adaptive is design software that can handle para-
metric designs. These allow a user to deine generalizable geometry that is
adjustable according to a few parameters. For example, a family of hammer-like
shapes could all be described using a single geometry with various lengths
and widths of their handles.
At this point, we move into the future realm. The following types of design
languages are highly experimental, found mostly in research labs and other
cutting-edge design experiments. In one approach called “design as program-
ming,” a computer describes a shape as a sequence of steps in a particular
order, somewhat like describing a cake by its recipe, not by its inal appearance.
Next, a more complex approach is offered by what we call “generative sys-
tems.” Such systems literally “grow” a shape from a seed, according to a given
set of rules.
