Dear list.
I've some idea i whant to share about GUI and pure-data. Seg_0.01 is the way i've chose to express them.
GOALS : To allow users of pure-data to disgn their own gui objects. PARADIGME : The gui will be entirely describe in pd patchs. One graphical object <=> One object in a pd patch. HOW : By provided an atom object called seg that can : * draw a line into a canvas, * allow to modifiy the line, * receive from the canvas and send to pure data all the events in relations with the graphical line object. * Destroy the line.
?
The goals of the experimental data features of pure-data is certainly the same as seg : Provide a tools to compositor to create their own representation for their own musical language.
Pure data will be THE PERFECT TOOL for writting music and not only to generate RT sound.
The data approch is based on the Concept of template : The user will disign the template of an object. The HUGE good point is that the user got the ablility to dynamicly cut and paste any objects. pure-data will automaticaly manage the duplication of the data. The bad point is that today the code that will manage the object is not inside the template. This mean that the way to write patch that use the data feature is not very easy.
Miller, please corect me if i'm wrong.
The seg approch is base on his paradigm : One pd object <=> one graphical object. This imply that the pd object will provide properties, methods and events of the graphical side. And this make very easy to make complex GUI with very few simples objects. But it's not possible to duplicate a graphical object without duplicate the pd object...
You can download seg_0.01 @ http://www.multimania.com/dh7/ I've put this link to the pdwiki too.
Any kind of feedback is welcome.
Damien Henry.
P.S. : Here the complete readme.txt file :
This is the README file for seg_0.01, a GUI (Graphical User Interface) server for pure data.
The goal of seg0.01 is not to replace the existing GUI, but to add some graphicals functionality to pd.
seg_0.01 is designed to run on all platforms supported by pure data. It is working very well on linux, well on win2000 and not so well on win98 where tcl/tk is too slow.
All this documentation is write in a poor english.
GOALS : To allow users of pure-data to disgn their own gui objects. PARADIGME : The gui will be entirely describe in pd patchs. One graphical object <=> One object in a pd patch. HOW : By provided an atom object called seg that can : * draw a line into a canvas, * allow to modifiy the line, * receive from the canvas and send to pure data all the events in relations with the graphical line object. * Destroy the line.
(linux) execute ./seg_0.001.sh from a terminal (windows) modify seg0.01.bat (put the rigth paths) and execute it. This will open a blank window. This window will be our drawing area.
Run pd and load the 00_connect.pd patch.
connect the patch with the upper left message
load one of the xx_test_seg.pd patch
Then play with the patch (uses the add messages one time to create some
gui) and enjoy the result in the seg window.
On a technical point view all of this is very simple.
seg is divided into 2 part : * The seg server (seg_0.01.sh) is a tcl/tk script that exchange data with pd via socket
* Some pd abstraction that are uses to establish the conection between pd
and the seg server.(s-link.pd) The "seg.pd" abstraction that is the basic object to draw a GUI in the canvas.
Because in iem or grip, the basic elements for buildings a GUI are on a very hight level : Sliders, and so much more. Seg didn't provide this king of high level features but allow the pd user to disgn their own ones.
The goals of the experimental data features of pure-data is certainly the same as seg : Provide a tools to compositor to create their own representation for their own musical language.
Pure data will be THE PERFECT TOOL for writting music and not only to generate RT sound.
The data approch is based on the Concept of template : The user will disign the template of an object. The HUGE good point is that the user got the ablility to dynamicly cut and paste any objects. pure-data will automaticaly manage the duplication of the data. The bad point is that today the code that will manage the object is not inside the template. This mean that the way to write patch that use the data feature is not very easy.
Miller, please corect me if i'm wrong.
The seg approch is base on his paradigm : One pd object <=> one graphical object. This imply that the pd object will provide properties, methods and events of the graphical side. And this make very easy to make complex GUI with very few simples objects. But it's not possible to duplicate a graphical object without duplicate the pd object...
As you can see thoses two way of thinking are very complementary, and i hope that some day all of this will merge.
fonctionalitys but in a more rational way.
Please don't hesitate to contact me if you need some help or some info about seg. Don't hesitate to reports bugs too...
Damien HENRY : dh7@multimania.com
Except as otherwise noted, all files in the seg distribution are
Copyright (c) 2002 damien Henry and others.
For information on usage and redistribution, and for a DISCLAIMER OF ALL WARRANTIES, see the file, "LICENSE.txt," included in the seg distribution.