The powder-calibration widget built its ring list with CalculateXtalRings(cell),
which assumes a primitive lattice. The overload taking a space group exists
precisely because the fit pairs the innermost OBSERVED ring with the innermost
LISTED one, so a list opening with a reflection the symmetry forbids scales the
whole calibration by the ratio between that ring and the first real one.
rugnux_cli was moved to the new overload when it landed; the viewer was not. So
on any centred sample cell the viewer's calibration was scaled wrong - the exact
failure the overload was added to prevent.
The widget already takes the cell from the loaded dataset; it now takes the group
with it and uses both. Where the dataset has no group the old call stands, which
is the same P-lattice assumption as before and no worse.
Found by a whole-branch review, which saw the two call sites side by side where a
per-commit read of either could not.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
leonarski_fandClaude Opus 5
2456bbe7f6