The question was whether simplification kept transit meaning
A geographic map is the honest input for Davao Transit. It shows where a route runs, but it can become hard to read when several lines share an avenue or converge near a terminal. A schematic map offers a different promise: fewer bends and more space between lines, at the cost of moving them away from their real geography.
The experiment was therefore not just an exercise in drawing cleaner lines. It asked whether a diagram could simplify geometry while keeping route identity, stop order, direction, shared corridors, loops, and transfer meaning intact. The proposed angle set was 0°, 45°, and 90°.
What is in the checkout
The current repository contains the geographic inputs, not a schematic editor or a generated schematic bundle. The bus file contains directional route features, and the jeepney file contains route geometries. There is no checked-out schematic package, editor entry point, generated image, approval artifact, or test that proves the editor described in the old page is part of the current application.
| Evidence | What it establishes |
|---|---|
| bus-routes.geojson | Current geographic bus route input |
| jeepney-routes.geojson | Current geographic jeepney route input |
| Git history search | No reproducible schematic or Loom implementation in this checkout |
| Loom WASM | An external research direction, not a current Davao Transit dependency |
That evidence changes how this experiment should be described. It is not accurate to say that Davao Transit currently converts these files into approved octilinear diagrams. The historical notes describe that direction, but the artifacts needed to inspect the result are absent.
The useful boundary was automatic draft versus editorial correction
The reported workflow was an automated first pass followed by manual editing and validation. That boundary makes sense for this network. A program can reduce bends and propose spacing. It cannot safely decide that two nearby lines represent one corridor, that a loop should read clockwise, or that two distinct services can share a berth without confusing a rider.
The editor, if recovered, should be explained as a maintainer tool. It would let a person correct a proposed diagram while preserving source identity and stop relationships. That is a different product from the geographic MapLibre map, and it is not a replacement for the planner's topology.
The same distinction applies to Loom. The project notes describe Loom as a possible WebAssembly direction for interactive schematic work. No current package import, build step, or deployment path here proves that Davao Transit uses it. The safe statement is that Loom is an external reference for future research, informed by this experiment's constraints.
Before and after evidence is unavailable
There is no repository-owned before/after screenshot, editor recording, generated SVG, or schematic JSON artifact for this experiment. The geographic GeoJSON files can serve as the input when the work is reproduced. They cannot stand in for an output image that is not present.
If the experiment is reconstructed, publish one real example with the original geographic view beside the edited schematic view. Annotate the exact route, stop, or corridor that changed, then explain what remained invariant. A Mermaid pipeline would imply a working implementation and would not answer the visual question this experiment was meant to answer.