@Strider007 You can search "duet object model", it should be move.axes[0-1].machinePosittion -- You can also explore the object model list in DWC

If it's something related to 'while expression', you can test the macro version like:

while true
  if sensors.gpIn[2].value != 0
    break
  if iterations > var.timeOut
    set var.tooLong = 1
    break

At least, I believe 'while true' is well tested.

I printed wire spools.

Something not useful, but fun, I printed some wind turbines -- for me, this is the driven reason to buy a 3d printer.

I plan to print some stacking boxes to use corner spaces around furniture, but have not started yet.

@mikeabuilder I don't have multiple tools, so it's simple. I never thought we can use slicer for this. In my start script, it tries different area (4 side rects) with different clearances, and calls this macro at the first suitable place. I think you may need to find a larger area so that multiple head can purge at the same time?

var min_x = ...
...

var xc = (var.min_x + var.max_x) / 2
var yc = (var.min_y + var.max_y) / 2

var clearance = 40 ; start from 40 / 2
var rect_x1 = 0
var rect_x2 = 0
var rect_y1 = 0
var rect_y2 = 0

while var.clearance > 1
  if mod(iterations, 4) == 0
    set var.clearance = var.clearance / 2
  if var.clearance < 1
    break
  set var.rect_x1 = mod(iterations, 4) == 0 ? var.max_x + var.clearance : move.axes[0].min
  set var.rect_x2 = mod(iterations, 4) == 1 ? var.min_x - var.clearance : move.axes[1].max
  set var.rect_y1 = mod(iterations, 4) == 2 ? var.max_y + var.clearance : move.axes[2].min
  set var.rect_y2 = mod(iterations, 4) == 3 ? var.min_y - var.clearance : move.axes[2].max
  if var.rect_x1 + global.NOZZLE_DIAMETER > var.rect_x2
    continue
  if var.rect_y1 + global.NOZZLE_DIAMETER > var.rect_y2
    continue
  if (var.rect_x2 - var.rect_x1) * (var.rect_y2 - var.rect_y1) < var.purge_area
    continue
  ; Project (xc, yc) into the rect
  set var.xc = min(max(var.xc, var.rect_x1), var.rect_x2)
  set var.yc = min(max(var.yc, var.rect_y1), var.rect_y2)
  ; Shrink the rect to purge_area, close to (xc, yc) and close to a square
  var xl = max(sqrt(var.purge_area), var.purge_area / (var.rect_y2 - var.rect_y1))
  var yl = var.purge_area / var.xl
  set var.rect_x1 = max(min(var.xc - var.xl / 2, var.rect_x2 - var.xl), var.rect_x1)
  set var.rect_x2 = var.rect_x1 + var.xl
  set var.rect_y1 = max(min(var.yc - var.yl / 2, var.rect_y2 - var.yl), var.rect_y1)
  set var.rect_y2 = var.rect_y1 + var.yl
  echo "Purge in the rect " ^ {var.rect_x1, var.rect_x2, var.rect_y1, var.rect_y2}
  M98 P"0:/macros/purge_rect.g" A{var.rect_x1} B{var.rect_x2} C{var.rect_y1} D{var.rect_y2}
  break

if var.clearance < 1
  echo "no place to purge"

I'm using this macro within the start macro, with all states are ready and some global variables, it purges the nozzle and writes a rectangle -- I have not tried, but I think you can use a narrow rectangle to purge a line.

I have tried and found G1 {var.w_sym}100 does not work with var.w_sym = 'X' or 'Y', which looks natural to me.

; Purge to a rectangle, or print a rectangle.
;
; This assumes everything is ready and uses
;
;  global.DEFAULT_SPEED: the default travel speed
;  global.NOZZLE_DIAMETER: the nozzle diameter, like 0.4
;
; This uses first layer height d * 0.75, width d * 1.25, while d is the nozzle
; diameter, with an extrusion factor 0.85.
;
; param.A-D: min/max x and min/max y of the rectangle
; param.S: the speed of print, default to 1440 (24 mm/s)
var x1 = param.A
var x2 = param.B
var y1 = param.C
var y2 = param.D

var purge_speed = 1440
if exists(param.S)
  set var.purge_speed = param.S

var tip_h = global.NOZZLE_DIAMETER * 0.75
var tip_w = global.NOZZLE_DIAMETER * 1.25
var purge_ratio = var.tip_w * var.tip_h / pow(1.75 / 2, 2) / 3.1415926 * 0.85

; write line in the w (longer) direction
var w_sym = var.x2 - var.x1 >= var.y2 - var.y1 ? "X" : "Y"
var w1 = var.w_sym == "X" ? var.x1 : var.y1
var w2 = var.w_sym == "X" ? var.x2 : var.y2
var h1 = var.w_sym != "X" ? var.x1 : var.y1
var h2 = var.w_sym != "X" ? var.x2 : var.y2

if move.axes[2].machinePosition < 1
  G1 Z2
if var.w_sym == "X"
  G1 X{var.w1} Y{var.h1} F{global.DEFAULT_SPEED}
else
  G1 Y{var.w1} X{var.h1} F{global.DEFAULT_SPEED}
G1 Z{var.tip_h}
M204 P300; the max print acceleration

var h_num = max(floor((var.h2  - var.h1) / var.tip_w + 0.5), 1)
var hd = (var.h2 - var.h1) / var.h_num
var h_e = var.purge_ratio * (var.w2 - var.w1) / var.tip_w * var.hd
while true
  if iterations >= var.h_num
    break
  if var.w_sym == "X"
    G1 Y{var.h1 + var.hd * (iterations + 0.5)}
    G1 X{mod(iterations, 2) == 0 ? var.w2 : var.w1} E{var.h_e} F{var.purge_speed}
  else
    G1 X{var.h1 + var.hd * (iterations + 0.5)}
    G1 Y{mod(iterations, 2) == 0 ? var.w2 : var.w1} E{var.h_e} F{var.purge_speed}

G10 ; retract
G1 Z1

; a print usually starts with retract + move + unretract, so we do not unretract here.

@T3P3Tony Thanks, I read the document, it's mentioned at the bottom clearly and exactly as this.

Do we know the consequence of too high jerk/instantaneous speed change values? Like smoothed corner or lost steps?

This problem persists with a large/full mesh. It's acts fine if I reduce the max speed from F24000 to F6000.

IIUC, if we have a smoothed mesh, this could be moderated? Like I can use a script to expand the heightmap.csv from a coarse grid (11x11 in 500mmx500mm) to very fine grid, it may be better?

I found my 3d printer has some issues of moving esp. after a print and even within a print, I tried many ways and final found it's caused by the mesh compensation.

With a small adaptive mesh probing area, G1 reduces the speed due to z speed limit (I think?), but the acceleration is too high that it makes very high vibration and/or even lost steps. When the head moves over the area, it looks like stopped abruptly and then restarts.

If we disable the mesh (G29 S2), the head can move freely in high speed again.

Any ideas / workarounds?

Thanks both, this confirms my understanding and the example configuration is very helpful.

I saw some release notes that we will have job.file.customInfo, which could already have this info. But I'm not sure it's accessible in the start.g or not -- I'd wait for a stable release to try this.

Is there a way to pass temperature, material, print ranges etc from the slicer to the start.g?

With this, we can put as much common starting part into the start.g file as possible. Otherwise, we may have to use slicer specific start part.