Algorithmic Architecture by Kostas Terzidis
"While the future of computers appears to include a variety of possible roles, it is worth exploring these roles in the context provided by the question: 'Who designs?' If one takes the position that designing is not exclusively a human activity and that ideas exist independently of human beings, then it would be possible to design a computational mechanism which would associate those ideas" (61).
This chapter takes a freewheeling look at the role of technology in design by tackling both the design process and scientific method. At times comments like, "once a design idea is agreed upon, the architects move into the production phases" seem to miss the importance of existing, however convoluted, integration of design and production (read science) methods at all phases. Nonetheless, the larger argument towards a unified strategy to mediate scientific and design methodologies resonates quite loudly with the current state of architecture. I would modify the author's final statement seen above to not imply that the machine is in any way a designer. I find this mysticizes the determinism of computation. I would argue that the algorithmic machine allows the designer to work at a different scale: at the scale of the rule and the system rather than only at the scale of the result. The algorithmic machine extracts results from the rules and systems, and no matter how impossible they are to imagine relating to the ruleset, the human is still the sole point of design in the process.
mark keller_08: reading
mark keller_08: random tetrahedron fractal
RANDOM TETRAHEDRON FRACTAL
Nested self-similar powercopies are used here to create a fractal tetrahedron geometry. Shown below is the unit powercopy. Each instantiation can be controlled by dictating the three points along the original triangle that determine the tetrahedron. Here these occur at the lines' midpoints, but can be set to any value, or randomized.
A height parameter responds to the area of the base triangle, while a universal multiplier allows control of the entire structure's relative height (see gif at top).
The color of the tetrahedron is driven by a rule that relates the tetrahedron's normal vector to the Z and X axes. These two angle values drive the R and G value respectively. Permutations of controlled and randomized structures follow:
I ran into difficulty with the naming of new powercopies, and had to manually rename each new powercopy to avoid redundancy in the design table. The other problem was that a the powercopy seemed to accumulate more sub-versions of that powercopy as I placed them. Below you can see a powercopy placed with the extra sub-versions highlighted. I was not able to find an efficient way to place nested powercopies without this happening, and simply went back in and deleted them.
mark keller_07: unfolded powercopy tower
UNFOLDED POWERCOPY TOWER
The powercopy geometry is formed from an input of three points, forming a plane on which a circle is drawn and copied along the Z axis. The area of the surface defined by the three input points determines the distance the second circle is copied, and also its radius. The original circle's radius is related to the length of the line it's centerpoint lies on.
A line connecting the two circles can be adjusted to define the twist of the resulting broken hyperboloid. The final geometry and corresponding unfolded shape are as follows:
When the framework is modified, the powercopies and their unfolded surfaces adapt. The powercopy is also controlled by a universal parameter that relates to the height of geometry. As this multiplier is varied, all powercopies react (see top).
