Featherframe 0.2.8

What's Changed

• docker-compose for self-hosting on a NAS (W-866) by @wr in #235
• Tests: pin the collage's day so a run straddling midnight passes by @wr in #236
• BirdNET-Go webhook detection source (W-865) by @wr in #237
• Gould's whole folio, and Region goes live (W-702, slice 3 of 3) by @wr in #232
• Copy: render/generate, not draw by @wr in #238
• Gould folio: record the Kansas catalogue cross-check (W-702) by @wr in #239
• Docs: an AI collage is bought per redraw when the species change, not once a day by @wr in #240
• Survey Gould's Australia, Asia and Great Britain folios (W-869) by @wr in #244
• BirdNET-Go setup steps match its real UI (W-865) by @wr in #245
• BirdNET-Go's rule is a file to import (W-865) by @wr in #246
• Folio groundwork: per-volume plate numbers and per-volume margins (W-874) by @wr in #249
• Push secret lives in the URL, with a New URL button (W-865) by @wr in #248
• Export the folios as an open dataset: wr/historical-bird-plates (W-868) by @wr in #242
• Gould scans fall back to the open dataset's release when BHL fails (W-868) by @wr in #243
• Gould's Birds of Great Britain fills Europe's seven gaps (W-872) by @wr in #247
• Gould's Birds of Australia folio, and an Australia region (W-870) by @wr in #251
• Gould's Birds of Asia folio, and an Asia region (W-871) by @wr in #252
• Australia: pin 15 of the 24 doubtful plates (W-870) by @wr in #253
• Status LED: one addressable RGB pixel on GPIO39 says what the frame is doing (W-876) by @wr in #255
• Open dataset: Gould's Australia and Great Britain (W-875) by @wr in #254
• Export Gould's Birds of Asia to the open dataset (W-871) by @wr in #257

Full Changelog: v0.2.7...v0.2.8