Owner's decision: "we need to abandon the perl oracle, this was only useful when initially porting, but I suspect many of the remaining problems have been carried in from the perl (such as the weird scoring of outdoor and circulation space, which definitely needs fixing)". 39 supports that second clause. Every defect the section found is inherited, not introduced: the two-sided crinkliness gaussian that double-charges surplus daylight (39.14), quality as a product over a variable number of factors (39.18), value_supported priced as value_inside so a terrace was worth more per m2 than a room (39.19), and circulation returning 0.07 per unit cost (hxi). So parity with the oracle was never a safety net -- it was a commitment to reproduce those defects. Each of 39.14, 39.18 and 39.19 would have been a parity failure had parity ever been checked, and keeping the tests would have meant reverting the fixes or explaining the failures away. Removed: oracle.py, test_oracle.py, the two parity tests and their fixture machinery in test_dom_corpus.py, innerloop.OracleEvaluator with its use_native and urb_root plumbing, the same plumbing through driver, and fourteen experiments/ scripts that could only run against Perl. Several of those are cited in earlier DESIGN sections; the citations now point into git history, which is the honest state -- they had been unrunnable since the oracle root (/home/bruno/src/urb) stopped being present. run_search is superseded by run_search_scaled, which does the same job natively. Kept: dump_areas.pl/.py, which validate GEOMETRY against Urb (4.1) rather than fitness, and the prose in fitness_cmd.py and dom.py explaining why the .score/.fails formats are shaped as they are. Provenance is worth keeping; a dead code path is not. CLAUDE.md updated: fitness.py is the only evaluator, and "Urb did it this way" is no longer an argument that a constant is right. 39.16 is the standing counterweight in the other direction -- the crinkliness target WAS right and twice looked wrong only because the code reading it was misunderstood. Inheritance is neither evidence for nor against. 410 passed. The 69 removed cases account exactly: 64 parity (all skipped, since no oracle .score was ever committed), 4 in test_oracle.py, and the guard test 39.20 added as a stopgap. Closes homemaker-py-118. Files homemaker-py-bk9 for the re-baseline that 39.19 made necessary. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01MJ84Feep79Hhm3E4zZJmnB
5.9 KiB
Project Instructions for AI Agents
This file provides instructions and context for AI coding agents working on this project.
Beads Issue Tracker
This project uses bd (beads) for issue tracking. Run bd prime to see full workflow context and commands.
Quick Reference
bd ready # Find available work
bd show <id> # View issue details
bd update <id> --claim # Claim work
bd close <id> # Complete work
Rules
- Use
bdfor ALL task tracking — do NOT use TodoWrite, TaskCreate, or markdown TODO lists - Run
bd primefor detailed command reference and session close protocol - Use
bd rememberfor persistent knowledge — do NOT use MEMORY.md files
Room-code namespaces (DESIGN.md §39.4/§39.6)
Leaf types share a first character across three namespaces:
C/O/S— generic structural types (circulation / outside / sahn), uppercase, reserved. A programme code spelled exactly one of these is rejected at load.- programme room codes — lowercase, may start with any letter. The generic
tests match
C/O/Sexactly, socr1is a room, not circulation. usage:— every space declares its access-requirement class (living/kitchen/bedroom/toilet/utility/none), mandatory, no fallback (DESIGN.md §39.7). A code's spelling decides nothing:name:is free text,usage:drives behaviour. There is no first-character type test left anywhere in the codebase.
When adding or editing a programme, run
python experiments/audit_programme_config.py — it reports reserved-name
collisions, the usage class each code picks up, and per-room-spec satisfiability.
Session Completion
When ending a work session, you MUST complete ALL steps below. Work is NOT complete until git push succeeds.
MANDATORY WORKFLOW:
- File issues for remaining work - Create issues for anything that needs follow-up
- Run quality gates (if code changed) - Tests, linters, builds
- Update issue status - Close finished work, update in-progress items
- PUSH TO REMOTE - This is MANDATORY:
git pull --rebase bd dolt push git push git status # MUST show "up to date with origin" - Clean up - Clear stashes, prune remote branches
- Verify - All changes committed AND pushed
- Hand off - Provide context for next session
CRITICAL RULES:
- Work is NOT complete until
git pushsucceeds - NEVER stop before pushing - that leaves work stranded locally
- NEVER say "ready to push when you are" - YOU must push
- If push fails, resolve and retry until it succeeds
Build & Test
pip install -e .
pytest
Architecture Overview
homemaker-layout is a Python successor to the Perl Urb project. It represents a building as a binary slicing tree where leaves carry target dimensions from the programme and division ratios are solved bottom-up (inverting Urb's top-down approach). The evolutionary search explores topology, types, and adjacency only.
Key modules:
dom.py— read/write Urb.domYAML into aNodetreegeometry.py— faithful port of Urb's top-down geometryprogramme.py— parsepatterns.configspace requirementssolver.py— bottom-up ratio solve (scipy)shapecurve.py— Otten/Stockmeyer shape-curve DP: exact size/width/proportion feasibility for a frozen topology, any storey count (DESIGN.md §37.2/§37.4-§37.6); used asdriver._evaluate's NM warm-start/hard pre-filtercpsat.py— exact room-code-to-leaf labelling via OR-Tools CP-SAT for a fixed topology (DESIGN.md §37.7); replacesoperators._assign_adjacency_aware's greedy/beam room placement behindassign_solver="cpsat", and powers theoperators.mutate_reassignin-search repair operatorfitness.py— native Python fitness evaluator (replaces Perl oracle)fitness_cmd.py—homemaker-fitnessCLI entry pointcollapse_cmd.py—homemaker-collapseCLI: finish-time global cell→room collapse (94g)graph.py— leaf-adjacency graph for programme-driven fitness checksgenome.py— topology genome: base-floor tree + per-storey deltasoperators.py— high-locality mutation and subtree crossoverinnerloop.py— ratio optimisation inner loop (Nelder-Mead / CMA-ES)driver.py— memetic search outer loopevolve.py—homemaker-evolveCLI entry pointbubble.py— 3D bubble-diagram adjacency fitness-signal prototype (DESIGN.md §27,mi7); validated NULL, not wired intofitness.py— reference only, do not build on without a new formulation
Conventions & Patterns
Scoring .dom files
Use the native homemaker-fitness command. Like the old urb-fitness.pl, you
must cd to the directory containing the .dom file first — the tool
resolves patterns.config, costs.config, and writes .score/.fails
relative to cwd:
cd /home/bruno/src/homemaker-layout/examples/programme-house
homemaker-fitness cf0b8a77e8b2325f92a7e7d150184a55.dom
The score is written to <file>.dom.score and failures to <file>.dom.fails; the numeric score is also printed to stderr.
fitness.py is the only evaluator. The Perl oracle it was ported from —
oracle.py, urb-fitness.pl, and the parity tests against them — is gone
(DESIGN.md §39.21). Those parity tests had never actually run: no oracle
.score was ever committed, so on a clean checkout every case skipped, and the
only cases that ever executed compared the native scorer with itself (§39.20).
The corollary matters when reading the objective: a constant or a rule that
looks odd is not thereby validated by "Urb did it this way". Several
defects found in §39 were carried straight over from the Perl — see §39.19 on
value_supported, and homemaker-py-hxi on circulation, which the owner has
ruled needs fixing.