[Fluxus] seq and scene-graph related issues

[Kassen]

Dear list,

This morning I was experimenting with animating things from Fluxa's "seq".
This is potentially useful as frequency analysis can only get us so far with
regard to transients and we might like to have visual counter-rhythms to the
audio ones. Because (draw-foo) only lasts a frame I was experimenting with
building objects every beat (and sequencing transformations), then clearing
the scene-graph every 16th beat. As a strategy I believe this holds promise,
however there are issues;

(clear), when called from inside seq's function gives a error about being
unable to pop. IMHO this should work, at least from a usability point of
view, but fixing it is beyond me. Maybe (seq ) is outside of the
scene-graph's scope exactly because it (seq ) shouldn't be cleared by
(clear)? That should stay as it is, because one buffer for Fluxa and one for
Fluxus is great fun with Jack sending Fluxa's audio out to the Fluxus in and
to the speakers. Maybe this is inherently unfixable because of that?

I then tried to instead delete all children of primitive "1", the scene
graph's root; that'd do the trick too. For this I wrote a "(flush)"
function, reconstructed here;

(define (flush)
    (for-each
        (λ (child)
            (destroy child))
        (with-primitive 1 (get-children))))

This works and is probably closer to what I need too, however it also gets a
error that primitive 1 can't be found. That's clearly wrong; when primitive
1 can't be found there is something very weird going on and clearly it can
indeed be found as we can destroy all scene-graph objects in this way. As
this error seems harmless I can write a "lazy" error-handler around the
operation, one that doesn't even report it, but that doesn't sound like the
proper thing to do long-term.

Yours,
Kas.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.pawfal.org/pipermail/fluxus-pawfal.org/attachments/20110517/972a9357/attachment-0001.htm>