>>
>> > So if I understand you right, you wonder whether I would say that we
>>> could sidestep the automaton, machine, and see it only from the
>>> perspective of the relation between algorithm and action. This is in
>>> a way a question about the scope and media of live coding - at least
>>> it implies a decision there, which could be a in favour of a very
>>> wide area.
>>>
>>> I was asking in a slightly different direction. I notice that the
>>> difference between interactive art or live electronic music and live
>>> coding becomes unclear, and so does the difference between machine
>>> control (steering) and programming (coding).
>>
>>Where does this difference become unclear? Or do you mean that we need to
>>work more on defining it? I think we do, and it's something I suppose I'm
>>exploring a bit.
>
> Yes, I meant this. As soon as e.g. the modification of a simulation
> of a physical mechanism becomes a possible medium of live coding,
> there is simply the question where it begins to become something else
> (please don't misunderstand, I don't want to say that there is no
> mixture, or even continuum, it is just working toward a clear
> characterization)
>
>
>>I think of programming as the building of a new machine expressed within
>>another. The outer machine is a set of primitive rules. This seems to be
>>fairly distinct from steering to me, which implies the control of only
>>one, fixed machine.
>
> Agreed, recursivity would be a starting point for me, too. Now what
> about this: Assume you have a game where your first person movements
> depend on the environment, and you may change this environment with
> your actions. If you interpret this situation in terms of the
> environment as the machine in the machine, then you are building /
> modifying a new machine by your actions. But it is a standard
> situation for a game, so nothing special at all.
I would say that case is not so interesting because the landscape only has
meaning for the player - i.e. it's not a machine in itself.
Where it gets harder is if there is an AI bot which interprets the
environment, now you are modifying something which is changing the
behaviour of a machine, a bit more interesting. A possible distinction is
that it doesn't _define_ the bot's behaviour (it's analogous to it's set
of parameters). If it did, this would be livecoding for me.
> Then we have the choice: Live coding is something like a subjective
> decision to regard this action as such, or to find a more essential
> difference.
It seems that it's subjective what a programming language is, so it might
be hard to move further - but I think "total definition of something's
behaviour" points to something, maybe. The level of built in abstraction
makes this harder to pin down, and is a sort of continuum - in the same
way as livecoding from prewritten scripts or lots of predefined functions
seems to make the case weaker (to the point of a function called
"PlayMyLivecodingTrack();")
> One possibility is to characterize live coding by its
> distance from physical modeling, by its character of being involved
> with the formal, or axiomatic. So I think Swami Fadereu is right if
> he refers to the sundial and to interaction as an example to pose the
> same question. I agree that abstraction is an essential
> characteristic here. But of course I am curious to see other views.
I think a definition of a machine for a machine to interpret must be
axiomatic (almost by erm, definition?).
I'm getting confused now, I need to read more about these sundials...
cheers,
dave
Received on Wed Jan 09 2008 - 12:19:12 GMT