Skip to content

M10: the funnel and the PEP 545 milestone #16

Description

@tamnd

The funnel and the PEP 545 milestone. Human work, plus deterministic tooling.

Checklist

  • A person reads all of tier 1 against the English
  • Entries unfuzzied one at a time as they are agreed. There is no bulk command and there should not be one
  • Last-Translator updated per reviewed file
  • Reviewed strings submitted through Transifex by the reviewer, under the Documentation Contribution Agreement
  • ROADMAP.md rewritten off pydocvi report coverage. The current one says zero words are formally translated, over a repo holding a million words of machine output, and that gap is the thing to fix
  • README states plainly what is machine output, what has been reviewed, and what PEP 545 actually requires
  • pydocvi apply --branch 3.14 produces the 3.14 catalogs from the same memory, which is what makes the shared-memory claim real rather than asserted

Exit

bugs.po, tutorial/ and library/functions.po fully human-reviewed and submitted through Transifex. The coverage table is generated and true. No claim anywhere in either repository that machine output satisfies PEP 545.

Metadata

Metadata

Assignees

No one assigned

    Labels

    corpus-repoChanges that land in python-docs-vi-machine-translationdocsREADME, runbook, reportsmilestoneA milestone tracking issue, M0 through M10reviewRound-trip, judge, sampling, the human funnel

    Projects

    No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions