[Dev] Intgra protocol specification

[Jamie Bullock]

Dear Integra dev'ers,

Welcome to the new Integra dev list. This is a new list for discussion
relating to Integra development. It will be accessible via a public
forum-like interface on GMANE. 

Henrik Frisk and I are meeting for a development session relating to the
Integra 'backend' technologies at the beginning of March. I am hoping
that this can help catalyse the development of a working prototype that
demonstrates the complete Integra system. By this I mean that I would
like to be able to show how a web-based database query can lead to an
XML file that can be loaded into an audio environment as an Integra
module, and controlled by the Integra GUI and actually produce sound!

One concern I have at the moment is that we don't yet have a fixed
Integra protocol specification to work to. NoTaM produced the original
specification, which was uploaded in April last year, and since then it
has been questioned on a number of different levels. The most recent of
the discussions about this was at NoTaM in our December development
meeting, but there have also been discussions online as well as in
Vilnius.

I think we urgently need a working protocol document that clarifies
conclusively the current status of the protocol, so that we can proceed
with a few basic assumptions. Would it be possible for someone at NoTaM
to draw this up, maybe in time for the beginning of March? It needs to
be a simple one-page html document, like the existing one
(http://22.alcyone.zettai.net/integra-OSC/interface_specification.html),
but it does need to be definitive: "version 1.0". It could probably be
done by modifying the existing document. 

Issues that need to be reflected are:

1. Fake namespace. We have talked about the idea of the namespace being
more like a 'marker' system than a means of communication, i.e. the user
sees /sinus1/freq 440, but under the hood, something entirely different
may be happening i.e. a more efficient protocol is used.

2. Address mapping. Can we get rid of /addressmap or not? 

3. Busses. The idea of discarding direct and exclusive connections
between modules in favour of a broadcast/multicast approach has been
discussed in detail, but what is the final conclusion?

I think there are a few more that I have missed, and maybe some
conclusions were reached in Oslo after I left. At any rate, I thought I
would start the ball rolling so we can get a formal written
specification together as soon as possible.

best,

Jamie