Changelog

First Windows release, at full feature parity with Redlinr 1.1.3 for macOS. This is also the version you get today: the fixes and the language pass below landed in the weeks after launch without a version bump of their own, so this page groups them under the same 1.1.3 that's still on VERSION.

Compare & redline

• Inline redline (deletions struck through, insertions underlined) plus an aligned split view of Original vs Revised.
• Two-level diff: paragraphs are aligned first, then a word-level diff runs only inside changed paragraphs, so even large documents stay fast. A single very long changed line falls back to a clean delete+insert instead of freezing the compare.
• Change navigation: a floating "Change N of M" pager (N / P) and a change-summary sidebar, both with a persistent highlight that follows along in the redline.
• Moved-block detection: a relocated paragraph is shown once with a "Moved" badge instead of reading as an unrelated delete-and-insert pair.
• Plain-English headline summarizing the most significant changes, with paired before→after value chips for numbers, money, dates, and percentages.
• Ignore case and ignore whitespace toggles re-run the compare with relaxed matching, while the rendered text always shows the original.
• Automatic reload: Redlinr notices when an open document changes on disk and offers a one-click reload, via a banner, a header button, or Ctrl R.
• Export the redline as a single self-contained HTML file (Ctrl E).

Reading your documents

• Word (.docx) and OpenDocument (.odt) are read with a pure-Python OOXML/ODF reader (stdlib zipfile + ElementTree, no lxml, no python-docx).
• PDF text layers via pypdf, RTF via striprtf, HTML with the standard-library parser. Markdown and plain text are read directly.
• Every extractor emits one line per paragraph (the same shape the Mac's textutil/PDFKit produce), so the two-level diff behaves identically on both platforms.
• Legacy binary .doc files aren't readable on Windows. Redlinr explains and suggests saving as .docx instead (the Mac build leans on Apple's own converter, which Windows has no equivalent for).
• No OCR: a scanned, image-only PDF has no selectable text, so Redlinr reports that plainly rather than guessing.

Windows fit & finish

• Per-user install into %LOCALAPPDATA%\Programs\Redlinr: no UAC prompt, no admin rights.
• WebView2 rendering with a Mica backdrop, Windows 11 rounded corners, and a themed titlebar that follows the in-app appearance toggle (Light / Dark / System).
• Taskbar icon with a proper AppUserModelID, and "Show in File Explorer" for exports.
• Self-update in place: Redlinr downloads the new installer, verifies its Authenticode signature and that its version matches what was advertised, installs over itself and reopens. Any check that fails deletes the download and sends you to the download page instead.
• Business license, and a License / Updates / Help / Review / Tip footer matching the rest of the twoplus11 apps.

Language & accessibility

• 13 languages: English, Spanish, French, German, Italian, Portuguese (Brasil), Dutch, Polish, Russian, Turkish, Japanese, Simplified Chinese, and Korean. Redlinr follows the Windows UI-language list by default, and the toolbar picker overrides it and is remembered, including the pure-Python reader's own error strings. The picker shows language tags (EN, PT-BR, ZH) rather than flags, because Windows renders flag emoji as bare letters.
• Correct typeface. An early build asked for Apple's SF Pro first, which doesn't exist on Windows and fell back to a generic sans; it's Segoe UI Variable throughout now.
• WCAG 2.1 AA, measured. Fixed three real contrast failures (secondary labels on inset surfaces, the boundary of text fields, the unchecked compare options, and the secondary confirm button), added full Windows High Contrast support, and made the toolbar reflow instead of clipping at 150% and 200% text scaling. Error toasts now interrupt properly instead of queueing.
• Corner radii come off a named scale, so the rounding is consistent from the window down to the smallest chip.

Known in this beta

• No OCR for scanned or image-only PDFs, which are flagged plainly rather than guessed at. The Mac app can read these with on-device text recognition; that hasn't been ported to Windows yet.
• A few things the Mac app has picked up since this port's 1.1.3 baseline aren't in this Windows build yet: a live-editable split view, a color-blind-friendly redline palette, and a sortable, exportable (CSV/Markdown/text) change list.
• SmartScreen may still warn on first download despite the signature; reputation builds per file. See the install notes.
• No screen-reader pass with Narrator or NVDA has been confirmed by a human yet. The accessibility work is tested structurally and checked with axe-core, but assistive technology hasn't driven the app end to end. If you use one, that feedback is the most valuable thing you could send.