I was talking about Gisele Beiguelman's didactic presentation of code errors a few weeks ago. Today I came accross one on Youtube, a perfect rectangle of code as a screen of noise worth mi_ga's ascii carpets.
Currently I am reading (and struggling with) Claude Shannon's Mathematical Theory of Communication. At some point he describes channels of communication with noise, and suggests a system of noise reduction. In order to ameliorate the system, an "observer" is posted so he can see both the message as emitted and as received, as well as the mistakes carried along. It writes down the errors and sends them trough a "correction canal" as data ready for correction ("données à corriger"). These mistakes, under certain conditions, can be processed and encoded and sent along with the message next time it is emitted, so as to reduce the purcentage of noise.
I think it is funny to consider the network and/or software user as involved in such a system of correction. This might work as a metaphor that would transpose this system into a moral context, thus helping us interpreting the observers as a kind of tattlers. In everyday use, we happen to relay the encoding of errors in software systems, willfully or not: the messages sent, from the basic 'page not found' 404 to pop-up window announcing the crash of your application, automated messages ask for or automatically trigger the report of mistakes to webmasters or maintenance robots. Denounce for the better, for the sake of the community! Be it laziness or a sense of solidarity between humans and robots, but these mistakes are rarely reported when the observer is is asked to and not forced to do it (personally i never denounce criminal robots!).
I think it is funny to consider the network and/or software user as involved in such a system of correction. This might work as a metaphor that would transpose this system into a moral context, thus helping us interpreting the observers as a kind of tattlers. In everyday use, we happen to relay the encoding of errors in software systems, willfully or not: the messages sent, from the basic 'page not found' 404 to pop-up window announcing the crash of your application, automated messages ask for or automatically trigger the report of mistakes to webmasters or maintenance robots. Denounce for the better, for the sake of the community! Be it laziness or a sense of solidarity between humans and robots, but these mistakes are rarely reported when the observer is is asked to and not forced to do it (personally i never denounce criminal robots!).
This triadic system questions the very use of information we are defining on a daily and forgetful basis. According to Shannon, the more the message is reencoded and duplicated, the less information it reveals. Sending error codes back to the webmaster (or the robot police) will replay the message through a more refined (noise-less) medium but lighten from its newness (its information). Of course in the case of Youtube, the message is replayed for a multitude of different receivers. But what when the receivers become the observers? When does new information arises?
What the community gains in efficiency, it looses in innovation: would that be the axiom of Shannon theory if it was intepreted in terms of social relationships (one reproach made to Shannon is that his theory does not take into account the symbolic side of communication)? I wonder how much the message is changed and if it allows to talk about new information, i.e. innovation. Shannon redundancy contradicts that. But what when these error codes and processes of reencoding actually help innovation - I am thinking about patches for free software. Tattling for the new!
Also I am being reminded of the interventions of Codeworkers on mailing-lists, which are particularly wicked because they stage error encoding (404 or error aesthetics, as they call it ) in order to break up the political redundancy of online communities. They make visible the invisible structure of system observation at play, and list-servs are indeed environments where tattlers abund (remember JODI's interventions on Eyebeam, and particularly the hilarious Cyberstar prank that I mentioned in a previous post). Mailing-lists have a set of rules (netiquette) that shape information into categories which tend to reflect the structure of their social behaviour (as included in the art milieu). Codework disturbances such as NN's or JODI's introduce a logical paradox: 1/ the Codeworkers, as net artists, need the communities if they want a space to perform into and a meaning to give to their performance (disturbing social entities), 2/ but these interventions weaken the body that they are a parasite of, or they even kill it (the history of the Syndicate mailing-list is exemplary).
What new pussycat? Are the codeworkers bringing new information themselves? This is an open question.
Also I am being reminded of the interventions of Codeworkers on mailing-lists, which are particularly wicked because they stage error encoding (404 or error aesthetics, as they call it ) in order to break up the political redundancy of online communities. They make visible the invisible structure of system observation at play, and list-servs are indeed environments where tattlers abund (remember JODI's interventions on Eyebeam, and particularly the hilarious Cyberstar prank that I mentioned in a previous post). Mailing-lists have a set of rules (netiquette) that shape information into categories which tend to reflect the structure of their social behaviour (as included in the art milieu). Codework disturbances such as NN's or JODI's introduce a logical paradox: 1/ the Codeworkers, as net artists, need the communities if they want a space to perform into and a meaning to give to their performance (disturbing social entities), 2/ but these interventions weaken the body that they are a parasite of, or they even kill it (the history of the Syndicate mailing-list is exemplary).
What new pussycat? Are the codeworkers bringing new information themselves? This is an open question.
Keywords: Claude Shannon, communication, correction, data, information, JODI, mailing-lists, message, mez, mi_ga, network, NN, noise, systems
Comments