Skip to content

v3

Choose a tag to compare

@karanshukla karanshukla released this 23 Oct 15:44
· 299 commits to main since this release

What's Changed

  • Six open issues: model tags, sandboxed document parsing, HEIC/AVIF, similar images, per-place switches, per-format counts by @karanshukla in #90
  • Test that iPhone portrait HEICs decode upright by @karanshukla in #91
  • settings: expose the window to screen readers, and test it headless by @karanshukla in #94
  • krunner: Read aloud with vinoSpeak, and Control1.RecentFiles / FileText by @karanshukla in #93
  • settings: follow the desktop's large-text setting by @karanshukla in #99
  • settings: egui 0.37, so screen readers hear disabled buttons as disabled by @karanshukla in #98
  • extract: read PDF fonts through ToUnicode before /Encoding by @karanshukla in #100
  • embed: hold each stage on the NPU to a stored CPU reference by @karanshukla in #102
  • extract: EPUB, subtitles, and Office/ODF macro and template variants by @karanshukla in #101

Full Changelog: v2...v3