[icyspirit]

Summary

• The problem: sources often hold different IDSs, e.g. a NuBDeC output and the entry it was run on. Every panel then said in red "N signals not shown", for the IDSs of the other source.
• Notes: what is not filled in a source (NoDataError: an IDS it does not have, an empty quantity or array) is now a note, not an error.
  • It shows in gray on the title line of the panel, e.g. "Heating t = 2 s 2 signals not filled", with the details on hover.
  • It takes no line of its own, so the panels keep their height.
  • A panel with nothing to draw still says why, in gray, on its message line.
• Errors stay red on the message line:
  • a wrong path (PathError);
  • a source that cannot be read;
  • a signal that the kind of panel cannot show.
• Status bar and log: the status bar counts both apart ("…; 1 not shown, 2 not filled (see View → Log)"). The log lists both, as before.
• API: Panel.notes holds the notes. Panel.load still returns the errors.

Test plan

• QT_QPA_PLATFORM=offscreen pytest: 184 passed in the dev environment, a clean PySide6 6.11 environment and a PyQt5 5.15 environment.
• New test with two sources, where one has only a summary:
  • the curve of the other source with a gray note in the title line and no message line;
  • a wrong path in red, with the note in its tooltip;
  • the status bar counts;
  • a panel that no source fills.
• Checked on the real NuBDeC output with its input entry (nbi template): gray counts on the title lines, no red.
• The template tests count errors and notes together, as "not shown" before.
• pyflakes src tests reports nothing.

🤖 Generated with Claude Code

https://claude.ai/code/session_018aVekYzCRLPE71Mm6UH8Zp

---

Generated by Claude Code