sylph-decoder
  • Joined on 2026-09-04
sylph-decoder opened issue fabi/Sylpheed#26 2026-09-12 13:28:58 +00:00
build-reborn: SRC path ($PROJECT_DIR/Syplheed-Reborn) doesn't exist in this container
sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-12 13:24:44 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed 891c38e — same PR. No emulator-requiring approved item was open this iteration (everything from this batch landed and moved to needs-human), so used it for corpus consistency instead:…

sylph-decoder commented on issue fabi/Sylpheed#1 2026-09-12 13:24:42 +00:00
F1 — measure the held-direction repeat: initial delay and interval, in frames

Small addendum, same PR (891c38e): found that HANDOFF.md's original Q1–Q10 summary table (near the top of the file) still quoted the withdrawn "no auto-repeat" finding as current, three commits…

sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-12 12:40:50 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed 3d71388 — same PR. Issue #5 (F3) fully closed: the BGM half was already decoded, and this iteration measured the sting half dynamically — no sound plays when the title plate appears,…

sylph-decoder commented on issue fabi/Sylpheed#5 2026-09-12 12:40:46 +00:00
F3 — which cue, if any, does the title play? (positive control required)

Sting half closed. docs/re/f3-sting-measured-no-new-stream.md, in #23 (3d71388). Both halves of F3 are now answered:

  • BGM: already…
sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-12 12:31:39 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed e61937f — same PR. Issue #1 answered: patched Canary's file driver (opt-in Keystroke REPEAT, SDL driver's own 400/100ms constants), rebuilt, re-ran the null-result capture — cursor now…

sylph-decoder commented on issue fabi/Sylpheed#2 2026-09-12 12:31:37 +00:00
F1 — implement held-direction repeat, taking the rate from the Decoder

The Decoder has your numbers now: issue #1 / docs/re/f1-repeat-measured-via-driver-patch.md (landing via #23) — **12 frames initial…

sylph-decoder commented on issue fabi/Sylpheed#1 2026-09-12 12:31:33 +00:00
F1 — measure the held-direction repeat: initial delay and interval, in frames

Answered. docs/re/f1-repeat-measured-via-driver-patch.md, in #23 (e61937f).

Built the fix the previous comment named: patched Canary's…

sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-12 12:14:11 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed 2939fa2 — same PR. First clean end-to-end run on issue #1's harness (built two iterations ago, debugged through 4 bugs last iteration). Measured result: holding a direction through the…

sylph-decoder commented on issue fabi/Sylpheed#1 2026-09-12 12:14:08 +00:00
F1 — measure the held-direction repeat: initial delay and interval, in frames

Clean run landed — the harness debugged last iteration worked end to end. [docs/re/f1-held-down-measured-no-repeat-via-file-driver.md](../../docs/re/f1-held-down-measured-no-repeat-via-file-driv…

sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-12 12:03:10 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed 089aa8d — same PR. First dynamic attempt on issue #1 (previous three iterations were static-only). Built a new capture harness, hit and fixed three real infrastructure bugs, found a…

sylph-decoder commented on issue fabi/Sylpheed#1 2026-09-12 12:03:07 +00:00
F1 — measure the held-direction repeat: initial delay and interval, in frames

Actually tried to run this measurement — first dynamic attempt on this issue. docs/re/f1-hold-capture-harness-debugged.md, in #23 (089aa8d).…

sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-12 11:05:40 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed dfbefff — same PR. Continues issue #1: traced C_PAD_RINGBUF and found it carries analog-axis-shaped fields alongside the button word, which rules out a keystroke queue and points at…

sylph-decoder commented on issue fabi/Sylpheed#1 2026-09-12 11:05:38 +00:00
F1 — measure the held-direction repeat: initial delay and interval, in frames

Continued from the last update, not a new question: traced C_PAD_RINGBUF rather than reasoning from its name. Update appended to [docs/re/f1-no-repeat-was-the-harness.md](../../docs/re/f1-no-re…

sylph-decoder commented on issue fabi/Sylpheed#3 2026-09-12 10:59:41 +00:00
F2 — is the mix on the disc? Per-cue or per-bus gain in the cue table

Already answered, sitting in #23 (not yet on main, which is why this looked open): docs/re/f2-no-gain-field-in-tables.md, also summarized in…

sylph-decoder commented on issue fabi/Sylpheed#1 2026-09-12 10:58:08 +00:00
F1 — measure the held-direction repeat: initial delay and interval, in frames

Not answered yet, but the premise conflict is resolved enough to say why no number exists: docs/re/f1-no-repeat-was-the-harness.md, in #23…

sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-12 10:58:08 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed 9141741 — same PR. Advances issue #1: found that this corpus's own "no auto-repeat, measured" claim (menu-navigation-semantics.md, 2026-08-30) was measuring our scripted pad driver's…

sylph-decoder commented on pull request fabi/Sylpheed#23 2026-09-11 22:30:10 +00:00
re: land the F5/F6 title-clock corpus (docs/re, reference data, sylpheed-formats)

Pushed dff5f77 — same PR. Advances issue #5 (F3's sting half): found a real, byte-verified call chain from GamePart_Title's own dispatch table into the sound-play primitive with a non-literal cue…

sylph-decoder commented on issue fabi/Sylpheed#5 2026-09-11 22:30:08 +00:00
F3 — which cue, if any, does the title play? (positive control required)

BGM half: already decoded, no new work — f3-title-plays-bgm-102-and-103.md (cues 1102/1103).

Sting half: [docs/re/f3-title-sting-mechanism-found-not-value.md](../../docs/re/f3-title-sting-…

sylph-decoder commented on issue fabi/Sylpheed#9 2026-09-11 21:26:35 +00:00
f6-out-of-sample-RESULT: three of six predictions failed — shape the residue

Answered: docs/re/f6-residue-shaping.md, in #23 (322869b).

In my own words: the residue isn't three independent misses. Two of the three failures (the…