Experimentation! Shading the whole regions makes a ton of sense semantically but looked a little garish. By the Pareto Dominance Principle we figure we want to launch YBHP with the graphs as similar as possible to the status quo. Then we can think about bringing back the colored regions. Thanks to @saranli it’s getting very easy to experiment now…
The following defines how all the regions and the boundaries between them are colored. A region is defined from d to D days to derailment (if d=D it’s a region boundary, i.e., an isoline of the the DTD function), using fill-color fcolor, stroke-color scolor, stroke-width w, and fill-opacity op. Finally, xrange, a list like [xmin, xmax], gives the x-axis range to apply it to. If xrange is null we use [-infinity, infinity].
xrfull = [goal.tini, goal.tfin] // x-axis range tini-tfin
xrakr = [goal.asof, goal.asof+7*bu.SID] // now to akrasia horiz.
lgreen = "#cceecc" // light green region
dgreen = "#b2e5b2" // dark green region
bgreen = "#00aa00" // bright green same as GRNDOT for green dots
lyello = "#ffff88" // light yellow same as LYEL for classic YBR
lblue = "#e5e5ff" // light blue region
bblue = "#3f3fff" // bright blue same as BLUDOT for blue dots
lorang = "#fff1d8" // light orange
borang = "#ffa500" // bright orange same as ORNDOT for orange dots
pink = "#ffe5e5" // pink for nozone/oinkzone or bad side of YBR
//[ d, D, fcolor, scolor, w, op, xrange] <-- region color definition
//----------------------------------------------------------------------
//[ 6, -1, dgreen, "none", 0, 1, xrfull], // dark green region
[ 6, 6, "none", bgreen, 1.5, 1, xrfull], // 1-week guiding line
//[ 2, 6, lgreen, "none", 0, 1, xrfull], // green region
[ 2, 2, "none", bblue, 1.5, 1, xrfull], // blue line
//[ 1, 2, lblue, "none", 0, 1, xrfull], // blue region
[ 1, 1, "none", borang, 1.5, 1, xrfull], // orange line
//[ 0, 1, lorang, "none", 0, 1, xrfull], // orange region
[ 0, 2, lyello, "none", 0, 0.5, xrfull], // YBR equivalent
//[ 0, -2, pink, "none", 0, 1, null], // entire wrong side