Context
This is a prepared-substrate specialization of one-dimensional fold-and-punch computation. Input bits determine which preassigned sites are open; the fold program itself is fixed before the input is known.
Definitions
- The address map sends a material point to its final coordinate, and is the complete layer fibre above .
- counts potential input-dependent sites, maximum layer depth, effective variable-bearing fibres, folds, material length, and minimum site separation.
Hypotheses and scope
- Clean alignment (GP): two site images coincide or are separated by more than twice the punch radius.
- No crease through a punch (NC): every punch interval lies within one affine piece of the address map.
- Interiority (IN): every punch image lies strictly inside the final footprint.
- Uniform overlay (UF): the set of maximal affine pieces covering a point is constant across each punch image.
- The punch radius satisfies .
Proof or evidence
The source memo gives the complete syntax, semantics, resource coordinates, and geometric assumptions. A later re-audit found that allowing zero-radius punches makes the intended through-hole semantics fail, so positive radius is now part of the definition.
Verification notes
The current statement was compared with the corrected theorem memo and the independent internal re-audit. The re-audit also corrected the meaning of uniform overlay: covering pieces remain constant, not the material points in the moving fibre.
Limitations
- This model does not describe paper folding in general.
- The feature sites are prepared before the fold program; no new punch or cut is made during computation.
- The present results use all-layers folds and one final Boolean observation.
Notes
The doctrine is a prepared-substrate specialization of the fold-and-punch machine of An, Demaine, Demaine and Ku (CCCG 2017), including its 1.5D narrow-strip variant. Punch is native to that model; what is non-native is that the punches arrive before the program rather than being produced by it. R>0 was added after an independent re-audit found the theorem false at R=0 with all other hypotheses holding vacuously.