For other reasons, I have been verifying the patches sent to Pgenerator and noted something that looks strange to me.
I have checked with the buit-in 24 patches grey ramp.
1) With no stabilisation I see 27 patches (the expected 24, plus 3 black identical patches at the end. One is issued after having acknowledged the end of the characterisation message. Maybe 2 patches in excess, but anyhow very minor issue.
2) With a no FS stabilisation, I see 52 patches (the expected 48 patches, plus, at the beginning, the last measured patch is resent, and, at the end, 3 black patches are issued. I guess it depends on some SW streamlining, but, anyhow negligible.
3) With a FS stabilisation, I see 100 patches (the expected 48 patches are doubled to 96, and the one at the beginning and the 3 at the end are still there).
Let me show the dump of these extra patches:
a) stabilisation patch:
FIRST
0000 e4 5f 01 de 65 3b 2c 3b 70 ec f6 7d 08 00 45 00 ._..e;,;p..}..E.
0010 00 66 81 39 40 00 80 06 f5 e8 c0 a8 01 15 c0 a8 .f.9@...........
0020 01 0a e1 16 00 55 82 5f d3 97 fb a2 9a f7 50 18 .....U._......P.
0030 10 01 76 6c 00 00 52 47 42 3d 52 45 43 54 41 4e ..vl..RGB=RECTAN
0040 47 4c 45 31 30 62 69 74 3b 31 39 32 30 2c 31 30 GLE10bit;1920,10
0050 38 30 3b 30 3b 34 34 39 2c 34 34 39 2c 34 34 39 80;0;449,449,449
0060 3b 30 2c 30 2c 30 3b 30 2c 30 2c 30 2c 30 3b 2d ;0,0,0;0,0,0,0;-
0070 31 02 0d 0a 1...
SECOND
0000 e4 5f 01 de 65 3b 2c 3b 70 ec f6 7d 08 00 45 00 ._..e;,;p..}..E.
0010 00 68 81 3b 40 00 80 06 f5 e4 c0 a8 01 15 c0 a8 .h.;@...........
0020 01 0a e1 16 00 55 82 5f d3 d5 fb a2 9a fb 50 18 .....U._......P.
0030 10 01 44 d2 00 00 52 47 42 3d 52 45 43 54 41 4e ..D...RGB=RECTAN
0040 47 4c 45 31 30 62 69 74 3b 35 37 36 2c 33 32 34 GLE10bit;576,324
0050 3b 30 3b 34 34 39 2c 34 34 39 2c 34 34 39 3b 30 ;0;449,449,449;0
0060 2c 30 2c 30 3b 30 2c 30 2c 36 37 32 2c 33 37 38 ,0,0;0,0,672,378
0070 3b 2d 31 02 0d 0a ;-1...
Basically, there is a first FS patch (449,449,449), followed, after the patch measurement, by a no FS stabilisation patch.
b) Patch:
FIRST
0000 e4 5f 01 de 65 3b 2c 3b 70 ec f6 7d 08 00 45 00 ._..e;,;p..}..E.
0010 00 66 81 cd 40 00 80 06 f5 54 c0 a8 01 15 c0 a8 .f..@....T......
0020 01 0a e1 16 00 55 82 5f eb 9e fb a2 9c 7f 50 18 .....U._......P.
0030 0f ff 4b d9 00 00 52 47 42 3d 52 45 43 54 41 4e ..K...RGB=RECTAN
0040 47 4c 45 31 30 62 69 74 3b 35 37 36 2c 33 32 34 GLE10bit;576,324
0050 3b 30 3b 36 34 2c 38 35 32 2c 36 34 3b 30 2c 30 ;0;64,852,64;0,0
0060 2c 30 3b 30 2c 30 2c 36 37 32 2c 33 37 38 3b 2d ,0;0,0,672,378;-
0070 31 02 0d 0a 1...
SECOND
0000 e4 5f 01 de 65 3b 2c 3b 70 ec f6 7d 08 00 45 00 ._..e;,;p..}..E.
0010 00 64 81 cf 40 00 80 06 f5 54 c0 a8 01 15 c0 a8 .d..@....T......
0020 01 0a e1 16 00 55 82 5f eb dc fb a2 9c 83 50 18 .....U._......P.
0030 0f ff 7c ef 00 00 52 47 42 3d 52 45 43 54 41 4e ..|...RGB=RECTAN
0040 47 4c 45 31 30 62 69 74 3b 31 39 32 30 2c 31 30 GLE10bit;1920,10
0050 38 30 3b 30 3b 36 34 2c 38 35 32 2c 36 34 3b 30 80;0;64,852,64;0
0060 2c 30 2c 30 3b 30 2c 30 2c 30 2c 30 3b 2d 31 02 ,0,0;0,0,0,0;-1.
0070 0d 0a ..
Basically, I see first the the right patch (64,852,64) and then a FS patch with still 64,852,64. From a timing perspective that second FS patch is sent close before the next FS stabilisation patch.
I cannot tell if it applies to other patch generators (Bug?) and if it comes from the compatibility need with other PG's (WIBNI?), anyhow unnecessary patches are sent.