Somewhat similar to this discussion, but more focused on live-coding
performances rather than studio production, is "Text" which was
presented at a local computer meetup recently by its author[1]. It's a
drag 'n' drop interface where component names and parameters are typed
in and connected together, similar to PureData, but the backend behaves
in a different way.

All of the text that is entered is Haskell code, eg. function names, and
the connections determine the (partial) order of the tokens. The really
bizarre part is that the connections are not specified by the user, they
are automatically determined based on the proximity of the components
(things which are close together will get connected, eg. by one becoming
a parameter to the other) and their types (components are only ever
connected in ways that are type-safe, so the program always* has correct
semantics).

The reasoning given for creating Text is that live-coding is very boring
to watch and there is little connection between what is seen (the source
code in a text editor) and what is heard. With Text, the flow of the
code and the meaning of the manipulations is more intuitive for the
audience. Also, the chaotic nature of the connections gives a lot of
unintended output which the coder can incorporate and follow.

Note that it's not just a music creation environment, as it can be used
to write any Haskell code (although it only seems useful for artistic
creations, rather than tasks with more rigid requirements).

Thanks,
Chris

[1] http://yaxu.org/text-update-and-source/

* Not sure how rigorous this is, it might just be a case of "works most
of the time"


_______________________________________________
fonc mailing list
fonc@vpri.org
http://vpri.org/mailman/listinfo/fonc

Reply via email to