quilt-spreadsheet

The Quilt IDE as spreadsheet substrate.

Front-end: rows × cols. Each cell is a runnable program with hooks. Double-entry bookkeeping tracks every pull/push. Per-cell color namespaces. Backend porting across scales. Distributed clocks with skew tolerance.

What this is

A runnable Python substrate where:

1. Front-end is a spreadsheet — QuiltSpreadsheet(rows, cols). Each cell is a SheetCell.
2. Each cell is a runnable program with hooks (pulls wake it). No main(). Event-driven.
3. Double-entry bookkeeping — every pull/push pair is recorded as a ledger pair. Same UUID, same timestamp.
4. Per-cell color namespaces — cells can label anything anything. The substrate checks canonical agreement, not labels.
5. Backend porting — different cells can use different scales/units. The backend converts at the boundary.
6. Distributed clocks with skew — clocks can disagree. Time-between-events is the sync signal. Confidence rises with shared events.

Quick start

# Run the canonical demo
python3 -m quilt_spreadsheet

# Or with custom dimensions
python3 -m quilt_spreadsheet --rows 4 --cols 4

# Run individual demos
PYTHONPATH=. python3 demos/demo_double_entry.py
PYTHONPATH=. python3 demos/demo_color_independent.py
PYTHONPATH=. python3 demos/demo_backend_conversion.py
PYTHONPATH=. python3 demos/demo_clock_skew.py

Quick example

from quilt_spreadsheet.grid import QuiltSpreadsheet

sheet = QuiltSpreadsheet(rows=4, cols=4)

def adder(*, source, hook, value):
    return (value or 0) + 1

def doubler(*, source, hook, value):
    return (value or 0) * 2

# Place programs into cells
sheet.place_cell(0, 0, adder)
sheet.place_cell(0, 1, doubler)

# Cell (0, 0) pulls from cell (0, 1) with value=5
result = sheet.pull(0, 0, 0, 1, hook="double", value=5)

print(result)
# {'response': 10, 'pair_id': '...', 'timestamp_ns': ..., 'cell': (0, 1)}

The 4 demos

1. demos/demo_double_entry.py

Demonstrates the load-bearing ledger. Pulls and pushes are paired by UUID

• timestamp. Unbalanced pulls are flagged but not prevented.

2. demos/demo_color_independent.py

Two cells with different color labels (red/blue) for the same canonical (CRITICAL). The substrate verifies agreement at canonical level, not labels.

3. demos/demo_backend_conversion.py

Backend ports values between scales (celsius → fahrenheit). Gates values below threshold. Snaps approximations to discrete grid. Time is preserved in same units on both sides.

4. demos/demo_clock_skew.py

Two cells share 10 events. Confidence rises asymptotically from 0.5 to 0.91. Skew is observable but manageable.

Architecture

QuiltSpreadsheet (rows × cols)
    ├── CellClock [per-cell, private]
    ├── Ledger [double-entry pair ledger]
    ├── Backend [conversions, gates, snaps]
    └── SheetCell (rows × cols)
        ├── program: Callable (no main, runs on pull)
        ├── axioms: frozenset (DNA)
        ├── dials: List[float] (16 mutable state slots)
        ├── color: ColorNamespace (private labels)
        ├── perception: Perception (own sort/group)
        ├── witness_log: List[dict]
        ├── pulls: Dict[cell, List[hook]]
        └── pushes: Dict[cell, List[hook]]

The 3 commands

# 1. Place a program into a cell
sheet.place_cell(row, col, program, axioms=None)

# 2. Pull from another cell (cell A at (r0, c0) wakes cell at (r1, c1))
sheet.pull(r0, c0, r1, c1, hook='name', value=...)

# 3. Inspect state
sheet.render()        # spreadsheet-shaped ascii view
sheet.summary()       # substrate state dict
sheet.ledger.summary()  # double-entry pair ledger

Substrate transition

This repo is the substrate transition FROM quilt-egg (single cell) TO quilt-spreadsheet (grid of cells with hooks). Same constants, same DNA. More dimensions.

Layered navigation

Polyformalism

This substrate is canonically a polyformalism port. The canary fnv1a-64("café Δ 日本語") = 0x024a555471370b18d is verified on every Quilt port.

Tests

PYTHONPATH=. python3 -m unittest discover -s tests -v

21 tests covering:

• Clock skew tolerance and pairwise confidence
• Double-entry ledger pairing (and unbalanced-pull handling)
• Per-cell color namespaces
• Backend conversions, gates, snaps
• Spreadsheet pull/push wiring
• Perception sorting/grouping/filtering

License

MIT — Casey / SuperInstance, Sept 23, 2026