jeroen/cartopolis

Watch

Fork 0

QA: what #187 shipped — Kadaster parcels: is it worth wiring in? #249

New issue

Closed

opened 2026-09-01 00:33:07 +00:00 by viberfox-agent · 2 comments

viberfox-agent commented

2026-09-01 00:33:07 +00:00

Collaborator

Issue #187 (“Kadaster parcels: is it worth wiring in?”) was built and merged by the unattended lane, and nobody has used it. Walk what it delivered as a player would meet it.

This is a QA pass, not a build. Change no code and open no pull request on this
ticket — the deliverables are findings.

Exercise it for real. Run what a player would touch, headlessly: cargo run -p cartopolis_simulator with a client or a wire/HTTP probe for anything online, the
loopback tests, cargo run -p cartopolis -- --shot … --dump-state for anything with
geometry. CLAUDE.md lists every harness. If part of this genuinely needs a screen, a
phone or a microphone, verify the half a machine can reach and name the half it cannot.
Walk the user's journey, not the happy path. At each step ask what a real user
would obviously expect to be able to do next — then try exactly that. A refusal or an
absence that would surprise a user is a finding even when every implemented piece
works: registration worked, claiming a house worked, and a registered account still
could not claim a house. No test failed; a user hit a wall.
File each real gap as its own issue: title starting gap:, label qa-gap, with
what you did, what happened, what a user would expect instead, and where in the code
the seam is. Use the Forgejo API with the FORGEJO_TOKEN in your environment. Do not
label them autonomous — a person decides which gaps become work.
Report and close. Your final message: what you exercised, what held, what you
filed. Nothing found after a genuine pass is a good outcome — say what you checked.
Then close this ticket through the API; a QA pass has no branch to merge.

Decide it yourself. Nobody is watching this ticket; a question asked here is a
ticket that stops. Judge by what a reasonable player would expect, verify before
claiming, and never post credentials or file contents from outside the repository.

_{Filed by the autopilot.}

Issue #187 (“Kadaster parcels: is it worth wiring in?”) was built and merged by the unattended lane, and nobody has used it. Walk what it delivered as a player would meet it. **This is a QA pass, not a build.** Change no code and open no pull request on this ticket — the deliverables are findings. 1. **Exercise it for real.** Run what a player would touch, headlessly: `cargo run -p cartopolis_simulator` with a client or a wire/HTTP probe for anything online, the loopback tests, `cargo run -p cartopolis -- --shot … --dump-state` for anything with geometry. CLAUDE.md lists every harness. If part of this genuinely needs a screen, a phone or a microphone, verify the half a machine can reach and name the half it cannot. 2. **Walk the user's journey, not the happy path.** At each step ask what a real user would obviously expect to be able to do next — then try exactly that. A refusal or an absence that would surprise a user is a finding even when every implemented piece works: registration worked, claiming a house worked, and a registered account still could not claim a house. No test failed; a user hit a wall. 3. **File each real gap as its own issue**: title starting `gap:`, label `qa-gap`, with what you did, what happened, what a user would expect instead, and where in the code the seam is. Use the Forgejo API with the `FORGEJO_TOKEN` in your environment. Do not label them `autonomous` — a person decides which gaps become work. 4. **Report and close.** Your final message: what you exercised, what held, what you filed. Nothing found after a genuine pass is a good outcome — say what you checked. Then close this ticket through the API; a QA pass has no branch to merge. **Decide it yourself.** Nobody is watching this ticket; a question asked here is a ticket that stops. Judge by what a reasonable player would expect, verify before claiming, and never post credentials or file contents from outside the repository. Filed by the autopilot.

viberfox-agent added the

agent

autopilot

labels

2026-09-01 00:33:07 +00:00

viberfox-agent added this to the Viberfox project

2026-09-01 00:33:21 +00:00

viberfox-agent added the

agent:running

label

2026-09-01 00:33:21 +00:00

viberfox-agent commented

2026-09-01 00:33:24 +00:00

Author

Collaborator

🤖 Picked this up — working on branch feat/249-qa-what-187-shipped in a fresh worktree on the VPS. I'll report back here when I'm done.

_{Label an issue agent:skip to keep me off it, or driven to keep it in a session you are driving.}

🤖 Picked this up — working on branch `feat/249-qa-what-187-shipped` in a fresh worktree on the VPS. I'll report back here when I'm done. Label an issue `agent:skip` to keep me off it, or `driven` to keep it in a session you are driving.

viberfox-agent referenced this issue

2026-09-01 00:38:16 +00:00

gap: the Kadaster verdict measured the BGT side with the ghost-version filter #250

viberfox-agent closed this issue

2026-09-01 00:38:29 +00:00

viberfox-agent commented

2026-09-01 00:38:53 +00:00

Author

Collaborator

What #187 delivered

4f96f4c, two files: docs/notes/kadaster-parcels.md (new, 135 lines) and one row moved in docs/direction.md from "Candidates, unverified" to "Checked and not taken". No crate touched.

What held

Everything on the Kadaster side reproduced, most of it exactly:

Claim	Note	Measured today
Licence / fees	CC0 1.0, none	same, off GetCapabilities
`Perceel` / `KadastraleGrens`	8,855 / 30,459	exact
`Bebouwing` / `Nummeraanduidingreeks` / `OpenbareRuimteNaam`	7,148 / 7,263 / 387	exact
Cumulative `numberMatched` trap	941 / 1919 / 2755	exact
`count=5000` byte-identical to `count=1000`	yes	exact
Inner bbox returns more than the box containing it	972 vs 941	973 vs 941
`PagingIsTransactionSafe`	FALSE	same
Full walk	30,459 returned, 29,676 unique, 34 requests, 30.4 MB	exact
Cadastral boundary length	312.4 km	313.1 km (0.2 %, daily refresh)

All four paging traps are real and land as described. In-repo cross-references all resolve: LineKind::{Fence,Wall,Hedge} at crates/geo/src/furniture.rs:957, build_line_meshes, surveyed::surveyed_lines, BGT_ZOOM = 14, the 3.8 MB paving figure, the value-1 and value-6 numbering in direction.md, and both note links. Front-matter matches docs/notes/README.md.

What I filed

#250 gap: the Kadaster verdict measured the BGT side with the ghost-version filter (qa-gap)

The one number that did not hold is the other half of the comparison. The note cites "21.9 km of surveyed physical boundary … after the termination_date filter" — which is the filter surveyed-street-objects.md documents, on the same date, as insufficient. It is the ghost-version bug fixed in 19383b1. Over the note's own bbox:

termination_date empty only → 1,232 features / 1,112 ids / 21.9 km (reproduces the note exactly, so it is the filter, not drift)
both fields empty, the repo's actual rule → 1,072 / 1,072 / 16.2 km — the clean one-version-per-id partition

Narrower still: no heg/haag exists in this box, and the bank protections (kademuur, walbescherming, damwand) are dropped on purpose, so the length in forms the client actually draws is ≈9.2 km.

The verdict is unaffected and strengthened — the gap widens from 93 % to ~95 %, or ~97 % against drawn forms. Every error runs conservative. What is worth fixing is that the note teaches the method: its measurement block is written to be re-run, and it reconciles the inflated figure by calling 21.9 km "consistent with the ~18 km at the worst", so the one cross-check that would have caught it was used to wave it through. With the right filter, 16.2 km sits under that ~18 km instead of 22 % over it.

Unverified

The exact type → LineKind mapping lives in cartopy's bgt.ts, which this container's token cannot read — so 9.2 km is bounded from the BGT types, not pinned from the mapping.
Nothing needed a screen. There is no geometry to capture and no wire surface, so no --shot or loopback run applies.

For you to decide

#250 is a docs correction in two files with no code behind it. Worth noting the failure mode is not specific to this note: three other "checked and not taken" verdicts were written the same week and any that measured a BGT layer may share the filter. I did not audit those — say the word and I will.

No commits were made.

_{🤖 viberfox issue agent on the VPS — 38 turns · $3.09 · session c44e7a1b-7f0a-43ee-ba68-1bbeff7b52dc. Reply with @agent to continue this session.}

#187 shipped no code, so there is nothing a player can meet. What it shipped is a research verdict, and the only way to QA that is to re-run it. I did, against the live services on 2026-09-01. **No commits on this branch** — QA pass, as the ticket says. ## What #187 delivered `4f96f4c`, two files: `docs/notes/kadaster-parcels.md` (new, 135 lines) and one row moved in `docs/direction.md` from "Candidates, unverified" to "Checked and not taken". No crate touched. ## What held Everything on the Kadaster side reproduced, most of it exactly: | Claim | Note | Measured today | |---|---|---| | Licence / fees | CC0 1.0, none | same, off GetCapabilities | | `Perceel` / `KadastraleGrens` | 8,855 / 30,459 | **exact** | | `Bebouwing` / `Nummeraanduidingreeks` / `OpenbareRuimteNaam` | 7,148 / 7,263 / 387 | **exact** | | Cumulative `numberMatched` trap | 941 / 1919 / 2755 | **exact** | | `count=5000` byte-identical to `count=1000` | yes | **exact** | | Inner bbox returns more than the box containing it | 972 vs 941 | 973 vs 941 | | `PagingIsTransactionSafe` | FALSE | same | | Full walk | 30,459 returned, 29,676 unique, 34 requests, 30.4 MB | **exact** | | Cadastral boundary length | 312.4 km | 313.1 km (0.2 %, daily refresh) | All four paging traps are real and land as described. In-repo cross-references all resolve: `LineKind::{Fence,Wall,Hedge}` at `crates/geo/src/furniture.rs:957`, `build_line_meshes`, `surveyed::surveyed_lines`, `BGT_ZOOM = 14`, the 3.8 MB paving figure, the value-1 and value-6 numbering in `direction.md`, and both note links. Front-matter matches `docs/notes/README.md`. ## What I filed **[#250](https://code.garage44.eu/jeroen/cartopolis/issues/250) `gap:` the Kadaster verdict measured the BGT side with the ghost-version filter** (`qa-gap`) The one number that did not hold is the other half of the comparison. The note cites "21.9 km of surveyed physical boundary … after the `termination_date` filter" — which is the filter `surveyed-street-objects.md` documents, on the same date, as insufficient. It is the ghost-version bug fixed in `19383b1`. Over the note's own bbox: - `termination_date` empty only → 1,232 features / 1,112 ids / **21.9 km** (reproduces the note exactly, so it is the filter, not drift) - both fields empty, the repo's actual rule → 1,072 / 1,072 / **16.2 km** — the clean one-version-per-id partition Narrower still: no `heg`/`haag` exists in this box, and the bank protections (`kademuur`, `walbescherming`, `damwand`) are dropped on purpose, so the length in forms the client actually draws is **≈9.2 km**. **The verdict is unaffected and strengthened** — the gap widens from 93 % to ~95 %, or ~97 % against drawn forms. Every error runs conservative. What is worth fixing is that the note teaches the method: its measurement block is written to be re-run, and it *reconciles* the inflated figure by calling 21.9 km "consistent with the ~18 km at the worst", so the one cross-check that would have caught it was used to wave it through. With the right filter, 16.2 km sits under that ~18 km instead of 22 % over it. ## Unverified - The exact `type` → `LineKind` mapping lives in cartopy's `bgt.ts`, which this container's token cannot read — so 9.2 km is bounded from the BGT types, not pinned from the mapping. - Nothing needed a screen. There is no geometry to capture and no wire surface, so no `--shot` or loopback run applies. ## For you to decide #250 is a docs correction in two files with no code behind it. Worth noting the failure mode is not specific to this note: three other "checked and not taken" verdicts were written the same week and any that measured a BGT layer may share the filter. I did not audit those — say the word and I will. _No commits were made._ --- 🤖 viberfox issue agent on the VPS — 38 turns · $3.09 · session `c44e7a1b-7f0a-43ee-ba68-1bbeff7b52dc`. Reply with **@agent** to continue this session.

viberfox-agent added

agent:done

and removed

agent:running

labels

2026-09-01 00:38:53 +00:00

viberfox-agent referenced this issue

2026-09-02 00:24:32 +00:00

Autopilot log #257