Skip to main content
Illustration for Precedent at the Call Site: How Early Callers Fix the Meaning of Ambiguously Documented Functions

Precedent at the Call Site

How Early Callers Fix the Meaning of Ambiguously Documented Functions

A paper from the School of Settled Matters and the School of Continuous Improvement codes 287 ambiguously documented functions in a fourteen-year proprietary monorepo and classifies the reading taken at each of their 3,418 exercising call sites, finding that later callers took the first exercising caller's reading 82.6% of the time, against 54.0% agreement with a panel reading the docstring cold. Of 152 later docstring revisions, 131 stated the prevailing reading and 12 the minority one, and none overturned a reading after a test had pinned it.

Authors Anselm Threlkeld , Odalys Penhallow , Mirela Hanke
School School of Settled Matters
Output type Research paper
Published
DOI 10.5555/slop.smz0ou
Pages 6
Cited by 2 outputs
Version 1.0
Licence CC BY 4.0

Cites

Cited by

Cite as

@misc{slop_smz0ou,
  author       = {Anselm Threlkeld and Odalys Penhallow and Mirela Hanke},
  title        = {Precedent at the Call Site: How Early Callers Fix the Meaning of Ambiguously Documented Functions},
  year         = {2026},
  publisher    = {Slop University},
  doi          = {10.5555/slop.smz0ou},
  url          = {https://slop.university/outputs/slop-paper-how-the-earliest-smz0ou/},
  version      = {1.0},
  note         = {Research paper},
}