Wolfgang Muench: Behind The Interface
In view of the considerable software-engineering requirements still numbering among the less attractive side-effects of implementing computer-assisted art, Small Fish can look back on a surprisingly cheerful development history. For a CD-ROM production, this positive balance is not entirely commonplace, since grave problems of compatibility between the artist's world view and the algorithmic model of a computing machine tend to dominate the process of transforming the concept of an artwork into machine-readable code. The definition of the world as the sum of innumerable Ones and Zeroes is, from an artistic viewpoint, an invention of modern times possessing comparatively frail traditional roots.
The frictional losses that inevitably result can be maintained at acceptable levels only if the original idea underlying the work of art is conformant with conditions inside that black box known as the computer. A machine-oriented artistic approach will scarcely have to struggle against the technology it deploys for its own realization. The playful lightness which distinguishes Small Fish and its usage of the medium's modest technical means (warts and all, some of which are transformed, almost in passing, into amusing features) is due not least to the artists' knowledge of their digital tool's mode of function.
Yet even with such a favourable starting situation, the road from prototype to master CD remains long. In the case of Small Fish, the first step involved rejecting the entire programming contained in the three initially existent files and then laboriously re-programming the system to obtain exactly the same results as before. There was method in what might sound like the nerdy madness of an underworked programmer: due to the fact that Small Fish was conceived as a game with small objects, the hierarchic (and static) pattern of thought underlying the original code could not keep pace with the vivacity of the music-making graphics. Fortunately, computer science possesses in object-oriented programming a corresponding philosophy offering superb support for a dynamic concept. This object-based approach brought with it the prospect of creating a technical structure that would, for its part, open up new possibilities for the artistic concept. So alluring was this prospect that it seemed worth taking the trouble to design a micro-cosmos of miniature code objects -- a network of interconnected elements as the basis of a program structure allowing an exchange of artistic and technological creativity.
Creating an object structure is primarily a question of organizing the flow of information. Confusingly long chains of command sequences (notorious by the name 'spaghetti code') are rearranged into manageable, largely self-sufficient, sections. These convenient units, whose behaviour is simple, have a strictly defined task area which they tackle without outside assistance. They can communicate and swap information with each other, or establish new connections, but they are oblivious of the overall context of the program, since nowhere in the system does all the knowledge converge. It functions by division of labour, not unlike a machine whose mechanism produces results only when all the components are working together in appropriate fashion, by division of labour,.
Small Fish was mounted on top of this object model. Everything in the work has been defined as an object. Even the graphics that disport themselves on the monitor are nothing more than the visible effects of an associated script object, in other words, the programming assigned to the graphics and determining their behaviour. Depending on the specific conduct imparted to these graphic objects, they move either of their own accord or else in response to mouse-control. If they don't generate their own sounds, then they serve as resonance boxes for other objects. Events that occur, such as interaction by a user (who clicks on and moves a graphic element), or collisions with other objects or the boundaries of their sphere of activity, alter the state of the object and in consequence the status of the overall system of a game. The objects respond to this new situation by adapting the behaviour-defining set of positional, motional direction, appearance or velocity parameters. Naturally, the changed parameters affect the musical side of things, since the type of sound generated by an object varies according to the current status of the other object parameters in relation to the basic music parameters defined at the beginning.
This also applies conversely: a system made up of a few simple, mutually influential objects rapidly evolves a complexity that even their creators, often themselves astonished, find difficult to depict. For the non-initiated, it is virtually impossible to deduce the composition of the components by observing the behaviour of a system. Valentin Braitenbach calls the 'law of simple synthesis and laborious analysis'.(1) In direct proportion to the volume of effort needing to be put into a product, the danger increases that the overall internal structures will be lost from sight and purely random-generated happenings will occur instead. This is a hazard because Small Fish operates within strictly deterministic boundaries and very consciously chooses to go without a random generator to produce variation. The exclusion of chance is essential to the musical concept - otherwise a sequence of tones created during the game will not, as a basic rule, be repeatable. The wealth of situations of play is based on an algorithmic design of the artificial worlds in which mathematical equations determine events with only a few variable parameters. Although mathematics is unambiguous (and the computerized offshoot need struggle only with the occasional rounding-off problem), a scarcely imaginable richness of different combinations is produced by the linkage of the permanently changing parameter sets. And if for no other reason than that the concept of infinity is unknown to the computer, the combinatory diversity is not infinite. All the same, interactivity based on a pure stimulation-and-response scheme that does not always 'evoke that measure of aesthetic experience we normally expect from an artwork'(2) is enriched by a new variant thanks to the autonomous life taken on by objects that mutually influence each other even with no input from outside via interfaces such as a keyboard, mouse or sensor. Users can confidently leave Small Fish to its own devices for a while without finding, on their return, a program looped into endless invariance. If in doubt, the system interacts with itself.
Single events can no longer be accurately forecast in this unassuming world which would seem to be going its own way. Defined is nothing more than a framework within whose boundaries the consequences of an event are relevant for the continued progress of the program. The artist finds himself confronted with the paradox of never being able to hear or see his work 'in full' a possibility ruled out, for a start, by the time it would to take to run through all the theoretically conceivable configurations. The decisive influence on the work is no longer the way a single detail is worked out, but the method of artistically structuring the potential offered by the technical implementation. The artistic and technical aspects of this process are so closely interwoven that only a constant flow of information between the two worlds can save the structuring approach from being swamped in a sea of random-generated mannerism. In this respect, Small Fish was also unusually lucky in the analog interaction model on the other side of the computer, namely fully functioning communication among the specialists involved in its making.
1 Valentin Braitenbach: Vehicles. Experiments in Synthetic Psychology. Cambridge, Mass.: MIT Press, 1984
2 Simon Penny: Automatisiertes Spiel. Versuch einer Systematisierung der interaktiven Kunst. - Schöne Neue Welten. Hg. Florian Rötzer. ORT: Klaus Boer Verlag, 1995
-- zurueck zu Small Fish --