/* stacked-names.css - one insider name per line, in both tables (09/08).
 *
 * WHY A FILE AND NOT THREE LINES IN style.css. Same arrangement as
 * congress.css, lifecycle.css, hover-chart.css, signal.css, row-click.css
 * and ai-summary.css: one file per feature, so "put the names back on one
 * line" is deleting this file and its <link>, not hunting declarations
 * through 34 KB of shared stylesheet.
 *
 * WHAT WAS WRONG WITH THE RUN-ON LINE. `insiders` is a TEXT column that
 * n8n joins with ", ", and the cell rendered it as one sentence: "Peters
 * Gregory K, BARTON RICHARD N, SARANDOS THEODORE A". Three facts read as
 * one string. Worse, the filings capitalise inconsistently - some names
 * arrive in full caps, some in title case - so the comma is the ONLY
 * boundary marker, and a comma inside a name ("SMITH JOHN JR, III") makes
 * even that boundary a guess. Stacking removes the ambiguity from the
 * rendering rather than from the data, which is where it actually lives.
 *
 * THE ROW GETS TALLER. That is the trade Carl accepted explicitly. It
 * buys back more than it costs: the column's width now follows the
 * LONGEST name rather than the sum of all of them, which is where the
 * horizontal room for the wider AI summary cell comes from.
 *
 * `display: block` on a <span> rather than a <ul>. A list would be the
 * more semantic markup and it is the wrong one here: <ul> inside a <td>
 * brings list-item semantics, default margins and a bullet reset to
 * fight, in a cell whose whole job is to be dense. The names are already
 * announced as a group by the column header.
 */

.insider-name {
  display: block;
}

/* Names are long and the column is narrow; without this a single long
 * name would either overflow or force the column wider than the stacking
 * was meant to make it. `overflow-wrap` breaks only where it must, so
 * ordinary names still break at the space. */
.insider-cell {
  overflow-wrap: anywhere;
  line-height: 1.45;
}

/* The link and the plain name must be distinguishable at a glance, and
 * this is the affordance that was missing on 09/08: nothing said WHICH
 * names had a page behind them. A name with no CIK in the window is not
 * an error and must not look like a disabled control - it is simply a
 * name we cannot resolve to a filer without guessing, which we never do.
 *
 * Underline on the link, nothing at all on the rest. Colour alone would
 * fail exactly the readers who most need the distinction. */
.insider-cell .insider-link {
  text-decoration: underline;
  text-underline-offset: 0.15em;
}