1.0.0-rc.174 (#84)

* Rugnux: Performance improvements on GPU and CPU (more of the pre-scan and of scaling on the GPU, faster CPU spot finding and crystal refinement), with unchanged results.
* Rugnux: More robust processing - patches of persistently hot pixels are masked, an inconsistent merge triggers a retry at the measured beam centre, and builds targeting different CPU levels give the same results.
* Rugnux: Improved scaling and merging - reflections with an overloaded pixel are dropped, as in XDS, sparse rotation sweeps are scaled more reliably, and French-Wilson amplitudes use an anisotropic Wilson prior.
* Rugnux: Improved space-group determination - glide planes in groups without a centre of symmetry, screw axes from short or weak axial rows kept when a higher group is adopted, and more reliable decisions on twinned and pseudo-symmetric crystals.
* Rugnux: Improved small-molecule processing - spots that grow wider than the integration disk and split spots are integrated over their measured footprint, sparse lattices are integrated on every frame, and the `.hkl` file holds unmerged scaled reflections (SHELX HKLF 4).
* Rugnux: Reads Rigaku d*TREK SMV images (Saturn CCD), including detector 2theta and encoded pixel overflows; home-source (rotating-anode) datasets were added to the validation battery.
* jfjoch_viewer: Fixed processing failing at the end with "Wrong JPEG library version" on Linux; the merge window shows the space group with proper subscripts and a checklist of crystal pathologies.

Reviewed-on: #84
Co-authored-by: Filip Leonarski <filip.leonarski@psi.ch>

tests/ModelScaleGPUTest.cpp

@@ -212,8 +212,12 @@ void CheckSameScale(const ModelScaleParams &gpu, const gemmi::Scaling<float> &cp
     const double b_cpu[6] = {cpu.b_star.u11, cpu.b_star.u22, cpu.b_star.u33,
                              cpu.b_star.u12, cpu.b_star.u13, cpu.b_star.u23};
     CHECK(std::fabs(gpu.k_overall - cpu.k_overall) <= 1e-4 * std::fabs(cpu.k_overall));
+    // b_star = F B_cart F^T, so a B of 0.001 A^2 is 0.001 / a^2 in it for the longest edge a. Without that
+    // floor a fit whose B comes out near zero is held to 1e-4 of almost nothing (a cubic case at B 0.6 A^2
+    // differed by 1e-4 A^2).
+    const double edge = std::max({cpu.cell.a, cpu.cell.b, cpu.cell.c});
     for (int i = 0; i < 6; i++)
-        CHECK(std::fabs(gpu.b_star[i] - b_cpu[i]) <= 1e-4 * MaxAbs(b_cpu) + 1e-12);
+        CHECK(std::fabs(gpu.b_star[i] - b_cpu[i]) <= 1e-4 * MaxAbs(b_cpu) + 1e-3 / (edge * edge));
 }
 
 struct Case {