Skip to content

v1.6 — AllSequences pointer table fix + FFI equivalence testing

Choose a tag to compare

View all tags
@github-actions github-actions released this 06 May 00:14
· 19 commits to main since this release
v1.6
ffcc1dc

AllSequences pointer table fix + FFI equivalence testing

Bug fixes

  • AllSequences and ThirtySequences pointer table was all zeros: The 60-entry pointer table at bytes 0–239 of the AllSequences SysEx payload was being emitted as 240 zero bytes. The SD-1 firmware and MAME use this table directly to locate per-sequence event data — they do not rebuild it from headers on receive. The table is now computed as cumulative 4-byte big-endian byte offsets into the event data area. For ThirtySequences, slots 30–59 all receive the total event data size as their value.

New features

  • sd1cli-ffi binary: New Rust binary that calls the sd1ffi C API functions directly (via rlib linkage) for all disk operations. Produces identical SysEx output to sd1cli, confirmed across 36 files covering all supported file types on a reference disk.

Testing

  • 135 unit/integration tests, all passing
  • New scripts/test-equivalence.sh: extracts every SysEx-capable file from a reference disk using both sd1cli and sd1cli-ffi, diffs outputs byte-for-byte; 36/36 pass
Assets 10
Loading