Skip to content

docs: record the considered-but-not-committed backlog - #25

Merged
jlc488 merged 1 commit into
mainfrom
docs/backlog-notes
Aug 8, 2026
Merged

docs: record the considered-but-not-committed backlog#25
jlc488 merged 1 commit into
mainfrom
docs/backlog-notes

Conversation

@jlc488

@jlc488 jlc488 commented Aug 8, 2026

Copy link
Copy Markdown
Contributor

Ideas weighed and set aside had nowhere to live, so the same discussion would get re-run later. docs/backlog.md keeps them with the reasoning, linked from both READMEs:

  • Candidate layouts — Persian (fa), Hindi InScript, Bulgarian БДС. Cheap (one defineLayout table each) and each brings its own audience; notes that fa shares the Arabic script range and bg shares Cyrillic, so both need the same coverage-based disambiguation ru/uk already use.
  • Deliberately NOT doing — the entry worth having: Turkish Q/F is Latin ↔ Latin, so toEn's script-range detection cannot tell which side a string is on. Adding it means an explicit-mode-only exception to the entire detection model, which costs more than the layout is worth. Without this written down, someone re-proposes it every year. (Chinese/Japanese are already excluded in the README for the candidate-selection reason.)
  • Extension follow-ups — options page (Latin target layout), contenteditable, _locales — held until v0.6 actually ships, since store feedback may reorder them.

It also records the signal the call rests on (0 issues ever, ~23 weekly downloads as of 2026-08-09) and the rule — demand-gated, an entry is not a promise.

Docs only; no version bump (0.5.2 was just published and there is nothing functional to republish for).

Ideas that were weighed and set aside had nowhere to live, so the same
discussion would just get re-run. docs/backlog.md keeps them with the
reasoning: candidate layouts (Persian, Hindi InScript, Bulgarian БДС —
cheap because each is one table, and each brings its own audience), the
extension follow-ups held until v0.6 actually ships, and — most usefully —
what we deliberately will NOT do and why.

The Turkish Q/F entry is the one worth writing down: it is Latin to Latin,
so toEn's script-range auto-detection cannot tell which side a string is
on. Adding it would mean an explicit-mode-only exception to the whole
detection model, which costs more than the layout is worth. Without that
note someone re-proposes it every year.

Also records the signal the call rests on (0 issues ever, ~23 weekly
downloads as of 2026-08-09) and the rule being applied.
@jlc488
jlc488 merged commit 12bb46d into main Aug 8, 2026
1 check passed
@jlc488
jlc488 deleted the docs/backlog-notes branch August 8, 2026 15:37
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