[PD] GOP drawing bug with scalars and z-order - test patch

[João Pais]

Following the discussion, here is a patch which should be able to 
document the issue. You'll need pd 0.5 and jmmmp 0.56 to try it out.

There is also a bonus bug related with loading a canvas from a file.

Best,

Joao
-------------- next part --------------
#N struct 1105-multiarray float x float y float color float thick float
x-scale float id float vis array 1105-multiarray-a 1105-ma-element
;
#N struct 1105-ma-element float y;
#N struct 1105-background float x float y float y1 float x2 float y2
float color;
#N struct 1105-ma-id float x float y;
#N struct 1107-color float x float y float farbe;
#N struct 1107-color-display float x float y float farbe;
#N struct 1110-color float x float y float farbe;
#N struct 1110-color-display float x float y float farbe;
#N struct 1113-color float x float y float farbe;
#N struct 1113-color-display float x float y float farbe;
#N struct 1117-multiarray float x float y float color float thick float
x-scale float id float vis array 1117-multiarray-a 1117-ma-element
;
#N struct 1117-ma-element float y;
#N struct 1117-background float x float y float y1 float x2 float y2
float color;
#N struct 1117-ma-id float x float y;
#N struct 1119-color float x float y float farbe;
#N struct 1119-color-display float x float y float farbe;
#N struct 1122-color float x float y float farbe;
#N struct 1122-color-display float x float y float farbe;
#N struct 1125-color float x float y float farbe;
#N struct 1125-color-display float x float y float farbe;
#N canvas 116 75 643 590 10;
#X declare -stdpath jmmmp;
#X obj 12 91 multiarray;
#X obj 12 71 r \$0-messages;
#N canvas 718 35 387 269 COPY 0;
#N canvas 0 50 450 250 (subpatch) 0;
#X array array1 100 float 3;
#A 0 0 0 0 0 0 -0.0285714 0.0428571 0.142857 0.2 0.242857 0.357143
0.378571 0.428571 0.428571 0.457143 0.457143 0.457143 0.428571 0.407143
0.385714 0.3 0.264286 0.2 0.1 0.0571429 -0.0142857 -0.0857143 -0.157143
-0.228571 -0.314286 -0.385714 -0.428571 -0.442857 -0.442857 -0.442857
-0.442857 -0.457143 -0.457143 -0.4 -0.371429 -0.3 -0.142857 -0.0857143
-0.0642857 -0.0285714 -0.00714285 0.0142857 0.0714286 0.0857143 0.1
0.142857 0.185714 0.228571 0.271429 0.314286 0.342857 0.378571 0.414286
0.485714 0.528571 0.628572 0.671429 0.671429 0.628572 0.585714 0.55
0.514286 0.428571 0.4 0.314286 0.257143 0.2 0.142857 0.0428572 0.0142858
-0.0714285 -0.114286 -0.2 -0.228571 -0.257143 -0.328571 -0.4 -0.442857
-0.471429 -0.485714 -0.485714 -0.485714 -0.478571 -0.471429 -0.464286
-0.457143 -0.428571 -0.428571 -0.414286 -0.385714 -0.371429 -0.371429
-0.342857 -0.314286 -0.157143;
#X coords 0 1 100 -1 100 70 1 0 0;
#X restore 18 167 graph;
#N canvas 0 50 450 250 (subpatch) 0;
#X array array2 70 float 3;
#A 0 0.771429 0.742857 0.685714 0.657143 0.6 0.571429 0.542857 0.514286
0.485714 0.457143 0.428571 0.4 0.371429 0.342857 0.328571 0.314286
0.285714 0.228571 0.171429 0.171429 0.171429 0.114286 0.0571429 0.0428572
0.0285715 -0.0285714 -0.0571428 -0.0714285 -0.0857142 -0.142857 -0.171429
-0.2 -0.228571 -0.228571 -0.228571 -0.257143 -0.285714 -0.314286 -0.314286
-0.342857 -0.342857 -0.357143 -0.371429 -0.428571 -0.438095 -0.447619
-0.485714 -0.485714 -0.485714 -0.542857 -0.571429 -0.571429 -0.6 -0.628571
-0.642857 -0.657143 -0.685714 -0.714286 -0.714286 -0.714286 -0.742857
-0.742857 -0.771429 -0.771429 -0.771429 -0.771429 -0.771429 -0.771429
-0.771429 -0.771429;
#X coords 0 1 70 -1 100 70 1 0 0;
#X restore 148 167 graph;
#X restore 536 125 pd COPY;
#X obj 237 125 s \$0-messages;
#X msg 237 94 all del \, create 1 \, 1 copy array1 \, create 2 900
2 \, 2 copy array2;
#X text 238 64 1 - This message cleans up multiarray \, adds and populates
2 new arrays;
#X text 238 264 3 - try moving the object in the left \, or minimizing
and maaximizing this window. Doing it causes a redrawing \, and the
same problem.;
#X obj 237 325 s \$0-messages;
#X text 237 194 The background is over the arrays. If you send a background
message \, the internal patch will redraw the arrays on top of the
backgrund (as a way of circumventing this problem).;
#N canvas 735 31 494 259 subpatch 0;
#X obj 14 90 multiarray 200 100 multiarray-example.txt;
#X obj 14 70 r \$0-messages;
#X msg 109 62 background 999;
#X text 109 19 Click here to fix the background. If you close this
subpatch and open it again \, the backgrround will still be over the
arrays.;
#X connect 1 0 0 0;
#X connect 2 0 0 0;
#X restore 238 241 pd subpatch;
#X obj 29 257 declare -stdpath jmmmp;
#X text 238 154 2 - the object inside this subpatch is the same as
on the left. But \, due to the reedrawing \, when you open the subpatch
the background is above the array scalars.;
#X msg 237 304 background 999;
#X msg 347 304 background 89;
#X text 13 520 Bonus bug: if you search for ## \, you'll find a section
in [multiarray] where a delay had to added. When loading a canvas from
a file \, there is a delay which isn't accounted for in the pd scheduling.
;
#X text 11 11 This patch explains the bug for the rendering of scalars.
To see this \, you'll need to install the jmmmp library 0.56 \, and
Pd 0.50. Follow the steps below:;
#X text 238 354 4 - bonus: if you create the object [pd-colors] (from
the jmmmp library) \, the GOP of multiarray will redraw. Probably this
is a know behaviour \, as new data structures are loaded into the system.
;
#X text 233 417 Conclusions: - usually scalars are drawn according
to their creation order \, as any other object \; - changing a scalar
causes it to redraw and come to the top \, which is also an expected
behaviour \; - what's usually not expected is that the drawing order
changes by other unknown reasons.;
#X connect 1 0 0 0;
#X connect 4 0 3 0;
#X connect 12 0 7 0;
#X connect 13 0 7 0;