kerfmaster pin list · Big capabilities
rulingA decision by Aristide or Jordan. True because it was decided; it can be superseded, but it cannot be stale.
Content last changed 2026-08-25 — computed from the item itself, not typed.
Ruling (Jordan, 2026-08-25, after build 134 snapped the ARCH VAC plate to its callouts): “Any time that a dimension is called out we'll need that to be accurate in the dxf — the only time a step/dxf file doesn't match 100% to the print is when the print rounds to 1 or 2 decimal places for tolerance purposes. For example, it is common to see .375" rounded to .38" when the tolerance key calls out +/-.005" for 3 decimal places and +/-.010" for 2 decimal places.”
What this binds: (1) A full-precision callout (3 decimals on these prints) is the design truth — delivered geometry hits it exactly; drawn-vector scatter inside the file never overrides it (the build-133→134 case: drawn 1.5038, callout 1.500, deliver 1.500). (2) A reduced-precision callout (.38) is the true value ROUNDED for tolerance purposes — never inflate it to .380 exact; the unrounded value comes from the drawn geometry or the obvious standard nominal (.375) inside the rounding band, and the decimal count picks the tolerance from the drawing's own key (+/-.005 at 3 places, +/-.010 at 2 on his example). (3) A drawn-vs-callout gap BEYOND the selected tolerance is a defect to report, not to silently snap. Applies to the future auto-snap step (pin drawing-pdf-extract — still unbuilt); build 134 already conforms (all its callouts were 3-place).