Skip to content

docs: describe the reference data by what it is - #367

Draft
luisleo526 wants to merge 1 commit into
mainfrom
lv/engine-prose-wording
Draft

luisleo526 wants to merge 1 commit into
mainfrom
lv/engine-prose-wording

Conversation

@luisleo526

@luisleo526 luisleo526 commented Oct 11, 2026 •

Copy link
Copy Markdown
Collaborator

This rewords comments, docs and READMEs that described where the reference trade lists and bars came from. They now describe the data by what it is. It changes no code, no test and no measured result.

What changed. 67 files: 42 Markdown files (fixture READMEs, docs, the root and benchmarks READMEs), 9 Python files and 16 C++ files (comments and docstrings only). Nothing under src/ or include/.

How it was checked.

  • Every changed Python file has an identical token stream and AST once comments and docstrings are removed; every changed C++ file has an identical token stream once comments are removed.
  • A read of the whole diff found it text-only and found no sentence that a tracked file does not support.
  • Two kinds of reference tape (the date-only one, whose last row is marked Open, and the unmarked one) are named consistently throughout.

What this leaves, and why.

  • Comment wording in six other files (.gitignore, CMakeLists.txt, one source file and three scripts) lands with a separate change, because they fall outside the paths a text-only change may take without a full sweep.
  • benchmarks/README.md:18 still carries an older phrase. The provenance checker (benchmarks/check_provenance.py) matches that exact phrase, so the README line and the checker's pattern change together in one small follow-up change that adds a test showing the checker still rejects a false claim.
  • benchmarks/results/selection.json and selection.md are generated output. The strings that produce them, and the probe identifiers they carry, change in a later change together with a regeneration, not by hand.
  • Identifiers that scripts and tests read are unchanged: an environment variable and its default path, a command-line flag, a field name in the population document, probe identifiers, string literals that tests compare, and the pinned values in the benchmark selection.
  • Fixture data files (metrics.json, meta.json and the like) are not touched; they belong to separate changes.
  • Wording of the form "exported" in prose that does not name a method is not changed here.

The diff view will show the removed lines, and they remain in history.

Reword comments, docs and READMEs that described where the reference trade lists and bars came from. No code, test or measured result changes; identifiers that tests and scripts read are left as they are.

The Pine v6 catalog steps from b1fbf0b now state only the result.
@luisleo526
luisleo526 force-pushed the lv/engine-prose-wording branch from 258784b to 54e63fa Compare October 11, 2026 04:48
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant