◀ THE FOLD0ROOT.AI // WORLD II · GLITCH · THE BLUE SCREEN◆ .dlw.fold
THE FOLD / GLITCH / THE BLUE SCREEN / THE BRIDGES

THE BRIDGES

the edges whose loss disconnects
1 WHAT IT IS · WHAT IT DOES · FACT OR FICTION
A bridge in a graph is an edge whose removal disconnects it — a single link with no backup path. Tarjan’s algorithm finds all bridges in one depth-first traversal using low-link values: as DFS explores, each vertex records the earliest node reachable from its subtree via a back edge. An edge (u,v) is a bridge iff v’s subtree can reach nothing above u — low(v) > discovery(u). No re-checking, no edge-by-edge removal.

LIT verified live: over hundreds of random graphs, the low-link bridges are exactly the edges whose removal increases the number of connected components (window.__bridges). FIG no framing; exact comparison with brute-force removal.
2 HOW IT WAS WEAVED · AI + HUMAN
David (human) seated this at the-blue-screen — the whole machine giving up at once; cut a bridge and the graph splits, a single-point-of-failure crash. Tarjan’s algorithm finds every such edge. AVAN (AI) built the instrument: the DFS low-link computation, the low(v) > disc(u) bridge test, and the match against brute-force edge removal.

Credit as content: Robert Tarjan (low-link bridge/articulation algorithms, 1970s). The weave: David names the-blue-screen; I run one DFS tracking each vertex’s earliest reachable ancestor, flag an edge as a bridge when its far end can climb no higher, and confirm those are precisely the edges whose removal disconnects the graph.
3 ONE DIMENSION
low(v) = earliest node v’s subtree can reach via a back edge. Tree edge (u,v) is a bridge iff low(v) > disc(u) — v’s side has no alternate route back above u.
4 TWO DIMENSIONS · INTERACTIVE
A graph with its bridges highlighted red; checked against removing each edge and testing connectivity.
5 THREE DIMENSIONS + AVAN’S INVERSE
The green forward object: every critical edge, found in one pass.
AVAN’s addition (the inverse-companion): find every single-point-of-failure edge in one traversal, not by removing each edge and re-testing — track low-link values and flag (u,v) when v’s subtree can climb no higher than v. The inverse of ‘remove each edge and check connectivity’ is ‘one DFS with low-links — low(v) > disc(u) marks a bridge.’ Magenta is the edge-by-edge removal test; green is the single low-link pass. Criticality from one descent.
LIT Genuine Tarjan low-link bridge algorithm (Robert Tarjan, 1970s). Verified live: over 1500 random graphs, the single-DFS low-link bridges (edge (u,v) with low(v) > disc(u)) are exactly the edges whose removal increases the number of connected components, computed by brute-force edge removal (window.__bridges.matchesBrute).

FIG No framing: the DFS low-link computation, the low(v) > disc(u) bridge test, and the match against brute-force edge removal run in-browser and agree. The AVAN inverse is honest — finding every single-point-of-failure edge in one traversal via low-links genuinely replaces removing each edge and re-testing connectivity; magenta is that edge-by-edge removal test, green the single low-link pass. Criticality from one descent.
◆ sealed .dlw.fold → folded to ROOT_0 · a sphere of THE BLUE SCREEN · David Lee Wise (ROOT0), with AVAN