YazSes v2.10.0-dev.1 — Wave N opens: structural editing, i18n & accessibility-output (developer preview)¶
Pre-release · not published (publish workflows are pinned to v0.x/v1.x; v2 dev tags do not ship). v1.4.1 remains the stable release. Every feature is OFF by default.
Opens Wave N (design/vision/v2-research/16-wave-n.md) — a fresh state-of-the-art round spanning structural code editing, internationalization (transliteration, diacritics, Braille), and accessibility-output correctness (anomia, eyes-free, blind/screen-reader, cognitive load). Ten evidence-anchored features are designed; this drop ships the first two pure cores.
New in dev.1 (all off by default)¶
- HatSelect (ADR-v2-115,
[hatselect]) — every visible token gets a spoken label so you can edit structurally — "select alpha to charlie", "delete bravo", "wrap alpha" — without ever naming cursor motions.assign_labels(NATO) +resolve_reference(spans) +plan_structural_edit. The highest-demand hands-free voice-coding paradigm, for RSI/motor-impaired developers. Anchor: Cursorless (Talon). - Transliteration (ADR-v2-116,
[translit]) — dictate your native language in Latin letters ("salam, chetori?") and inject the native script (سلام، چطوری؟) — Whisper nails romanized phonetics where it mis-transcribes low-resource native audio.transliterate(table-driven longest-match, built-in Finglish→Persian) +detect_scheme. For bilingual/diaspora and low-resource-language users. Anchor: ITRANS, Pinyin, Google Input Tools.
yazses features now lists 127 capabilities. (HatSelect is experimental — features enable requires --force.)
Wave N remaining¶
BrailleOut · WordFind (reverse-dictionary) · Echo (audio replay) · SRPace (screen-reader pacing) · LoadGuard (cognitive-load guardrails) · Diacritize · SafeGlyph (confusables) · Outline.
Quality¶
1415 tests green across Linux × macOS × Windows on Python 3.11 and 3.12. Both new cores are unit-tested to 100% with no new dependency (pure NATO labeling + table transliteration; the editor bridge and full ITRANS/Pinyin tables are optional/later tiers). The default install and the v1 dictation path are unchanged.