2010/8/13 Mathieu Bouchard matju@artengine.ca:
On Fri, 13 Aug 2010, Bernardo Barros wrote:
Hum, yes. For PD it is a problem indeed. The file format would have separate "content" from "presentation", right?
Yes and no... there are different things that this idea can mean. the documentation system of GridFlow does separate part of the presentation, as for example, the look of each section heading is completely determined by an abstraction, and the abstractions know the rules and know how to move themselves downwards on the canvas to leave enough space for other objects (you have to see it to understand) but the positions of the objects are still saved in the file even though most of them could be set to 0 0 when saving the file and the process of reloading the file would automatically reposition them (this latter repositioning already works).
I think if PD wants to have diff systems and things like that it have to somehow separate what is significant and not significant for the procedures of the program. Moving objects a little around is most of the times little relevant.
There is this dot (graphviz) language, do you know? It renders simple schemes to complete graphics. Something like:
one -> two -> three one -> four one -> five three -> six
Yeah, but this will add computations to PD to proceed just to show up a patch, maybe bad idea.
You can have the same program with objects in different places. But to have that you must guarantee that the object location cannot influence the procedures of the program (if a box is on the left or right of another one).
[inlet] [inlet~] [outlet] [outlet~] have behaviour that depends greatly on their x location.
Maybe a little change here? [inlet 1] [inlet 2] [inlet 3] ?