mirror of
https://github.com/jmcorgan/fips.git
synced 2026-10-05 19:18:25 +00:00
Johnathan Corgan
6b6d2a8df9
The stage block and the sections below it now share one left margin, the stage rows carry shorter text, and a blank line separates the stages by the question each answers: where is the target, how do we get there, can we reach it. The path row no longer appears when it computed a route. It restated the heading of the path section printed below it, and that section said the same thing in full. The row is kept when the stage failed or was skipped, where it carries a finding that appears nowhere else in the block. That exposed a defect the row had all along: its text ignored the stage's reason, so a disjoint_trees or no_next_hop failure printed the success wording. It now keys on the reason like every other stage. Also in the report: the npub moves onto its own line so the three identifiers align, the depth column no longer shifts with the length of the coordinate list, the route line sits directly under the two coordinate lists it derives from and is set off by a blank line on each side, the next-hop class loses its underscore, and the round-trip row names the exchange count only when it is not one. The path section heading, which said the route was computed rather than observed, is gone; the JSON keeps computed_locally and observed. Off a terminal the block is rebuilt from the settled prefix rather than appended row by row, because a group separator belongs to the row below it and a single-row render cannot know whether that row exists yet. A stage-pipeline diagram lands alongside the report, drawing the five stages left to right at 16:9 with each stage's failure reasons below it, and the bypass that skips both lookup stages when the coordinates are cached or the target is a direct peer. Its branches come from the probe state machine rather than from the report, so the path stage is drawn as the one failure that does not stop the probe.
