Found reviewing #49.
Four rules now outline a heatmap cell and they do not agree on where the ring sits.
td button:hover gained outline-offset: -2px, so hover draws inside the cell.
td button:focus-visible draws two inset bands.
td.hi draws a 2px outline with no offset, so it sits outside the cell.
td.sel does the same.
With border-spacing cut from 3px to 2px, the outward outlines of two adjacent highlighted cells each claim the full 2px gap between them, so the column highlight reads as one continuous band instead of nine rings. For a whole-column highlight that may be the better look, which is why this is filed rather than fixed. Worth a decision either way, so the four rules stop disagreeing by accident.
Separately, td.sel is dead. The class is declared in the stylesheet and never applied: state.sel in the graph code is unrelated and no code path adds sel to a heatmap cell. Searched the fragment source and the generated page.
🤖 Generated with Claude Code
https://claude.ai/code/session_01BXHrBTSU8fU4Zg7HfaMKs9
Found reviewing #49.
Four rules now outline a heatmap cell and they do not agree on where the ring sits.
td button:hovergainedoutline-offset: -2px, so hover draws inside the cell.td button:focus-visibledraws two inset bands.td.hidraws a 2px outline with no offset, so it sits outside the cell.td.seldoes the same.With
border-spacingcut from 3px to 2px, the outward outlines of two adjacent highlighted cells each claim the full 2px gap between them, so the column highlight reads as one continuous band instead of nine rings. For a whole-column highlight that may be the better look, which is why this is filed rather than fixed. Worth a decision either way, so the four rules stop disagreeing by accident.Separately,
td.selis dead. The class is declared in the stylesheet and never applied:state.selin the graph code is unrelated and no code path addsselto a heatmap cell. Searched the fragment source and the generated page.🤖 Generated with Claude Code
https://claude.ai/code/session_01BXHrBTSU8fU4Zg7HfaMKs9