- Disc::copy() hardcoded batch=64 sectors, exceeding BU40N's 60-sector
hw limit. Now accepts batch_sectors param, defaults to 60.
- IFO PGC: playback time at offset 0x04 not 0x02, cell time at cell+4
- DiscStream: set demuxer from content_format (TS for BD, PS for DVD)
- Flush TS/PS demuxers at EOF to avoid losing last PES frame
- M2tsStream: flush demuxer at EOF
- StdioStream: FMKV metadata header for roundtrip compatibility
Architecture:
- One stream per format, bidirectional PES (read/write on same type)
- IsoStream merged into DiscStream (one type, any SectorReader)
- Disc::copy() for disc→ISO raw sector dump
- IOStream trait deleted, all byte-level Read/Write removed
- ContentReader/OpenDisc/open_title/open_input/open_output deleted
- CountingStream wrapper for progress tracking
Error codes:
- All io::Error English strings replaced with Error enum variants
- From<Error> for io::Error conversion
- Unused variants removed, new stream/mux variants added
Deleted: mkvout.rs, pesout.rs, isowriter.rs, mkv-muxer-plan.md
Updated: all docs, README stream table, CHANGELOG
238 tests, 0 clippy warnings.
Streams are streams — they take impl Read, not Read+Seek or File.
- MkvStream::open takes impl Read (was Read+Seek)
- EBML element skipping uses skip_bytes() instead of seek(Current)
- Byte position tracking uses remaining-bytes counter, not stream_position()
- Removed file_size from open (progress is CLI concern)
- M2tsStream::open takes impl Read (was Read+Seek)
- Buffers first 1MB for FMKV header / PMT scan
- Uses chain reader (buffered head + rest) for sequential reading
- Duration unknown without seeking (0.0) — CLI can set from metadata
- Removed ReadSeek trait (no longer needed)
- WriteSeek kept (MKV muxer container format requires seeking internally)
Design fix: codec_privates are now a field on DiscTitle, not a separate
parameter passed through the pipeline. This eliminates the root cause of
the network codec_private bug (forgot to pass the separate param).
API changes:
- output() takes (url, &DiscTitle) — no separate codec_privates param
- MkvOutputStream::create, M2tsOutputStream::create, NetworkOutputStream::connect
all read codec_privates from title.codec_privates
- M2tsMeta::from_title() takes only &DiscTitle — reads privates from title
- Deleted from_title_with_privates (was the wrong-name duplicate)
- Merged read_header + read_header_from_stream into one read_header(impl Read)
- Deleted finish(self) from TsMuxer, keep only finish(&mut self)
Rule: ONE public method per action. No _with_X, _from_Y, _ref variants.
Bug 1: M2TS roundtrip dropped frames — TsMuxer converts length-prefixed
NALs to Annex B, prepends VPS/SPS/PPS from HEVCDecoderConfigurationRecord.
Bug 2: MKV remux lost codec_private — MkvStream.codec_private() now returns
data from EBML header.
FMKV header carries codec_private (base64) per video stream for lossless
M2TS roundtrip.
- M2tsOutputStream writes FMKV metadata header before TS data
- NetworkOutputStream sends FMKV header on connect
- Enables M2TS→MKV and network roundtrip to work correctly
- Remove IOStream, Read, Write impls from NetworkStream
- PES write sends FMKV header before first frame (protocol fix)
- PES finish sends TCP shutdown for clean EOF
- Update tests to use PES roundtrip instead of byte-level
- Remove NetworkStream from open_input/open_output (use input/output)
- Remove unused pes_buf field from M2tsStream and unused TS_PACKET/BD_TS_PACKET constants
- Replace match-with-single-pattern with if let (3 instances in drive/mod.rs)
- Replace match-can-be-? with ? operator for scsi::open call
- Add type aliases PesSetup and MkvHeaderResult to reduce type complexity
- Collapse identical if/else branches in tsmux.rs build_pes_header
- Use RangeInclusive::contains instead of manual range checks
- Make WriteSeek trait pub (was pub(crate) but leaked through pub fn)
- Remove empty line after doc comment in disc.rs
- Fix doc list item indentation in scsi/linux.rs (12 instances)
- DiscStream: BD (TsDemuxer) or DVD (PsDemuxer) auto-detected
- NetworkStream: Stream impl with PES serialize/deserialize
- StdioStream: Stream impl with PES serialize/deserialize
- MkvStream: Stream read returns PesFrame from EBML blocks
- M2tsStream: Stream read via TsDemuxReader
- PesFrame: serialize/deserialize for wire format
- TsDemuxReader: shared BD-TS demux helper
- All inputs and outputs support PES
- Unified Stream trait: read() and write() on one type
- PesFrame serialize/deserialize for wire format
- TsDemuxReader: shared BD-TS demux for any Read source
- MkvStream: PES read from MKV (EBML → PesFrame)
- M2tsStream: PES read via TsDemuxReader
- Network/Stdio output: PES serialization directly (no BD-TS wrap)
- Network/Stdio input: deferred (needs PES deserialization protocol)
- TsMuxer for M2TS output from PES frames
- input() and output() functions return Box<dyn Stream>
Stream trait: read() returns PesFrame, write() accepts PesFrame.
A stream is a stream — you read from it or write to it.
No separate Input/Output traits.
API: libfreemkv::input(url) and libfreemkv::output(url, title, codecs)
Returns Box<dyn Stream>.
- TsMuxer: PES frames → BD-TS packets (new, reverse of TsDemuxer)
- M2tsOutputStream: PES → TsMuxer → file
- NetworkOutputStream: PES → TsMuxer → TCP
- StdioOutputStream: PES frames → stdout
- NullOutputStream: discard
- MkvOutputStream: PES → MKV mux
- All outputs via open_pes_output()
- All inputs via open_pes_input() (ISO, disc)
- No byte-level fallback — everything is PES
- pes.rs: PesFrame, InputStream, OutputStream traits
- sector.rs: SectorReader now public, added FileSectorReader (ISO = file)
- Foundation for unified DiscStream that handles both disc and ISO
When the MKV muxer transitions from Scanning to Streaming phase,
it creates a fresh TsDemuxer. The old demuxer's remainder bytes
(partial 192-byte BD-TS packets) were lost, causing the new demuxer
to lose sync. All subsequent feed() calls found 0 packets.
Fix: transfer remainder via take_remainder/set_remainder so the
new demuxer maintains packet alignment.
Previously returned HandshakeResult with zeros when all host certs
failed. Now returns None. Also propagates volume_id read failure
instead of silently using zeros.
Previously used all-zeros AES key when unit_key_idx was out of range,
producing silently corrupted output. Now returns DecryptFailed error.
Also wired --raw flag through InputOptions → set_raw() on streams.
IsoStream decrypts in its read path using keys from scan.
IOStream trait gets keys() method with default DecryptKeys::None.
Pipeline no longer handles decrypt — just reads decrypted bytes.
- IOStream::keys() default returns DecryptKeys::None
- IsoStream and DiscStream override with real AACS/CSS keys
- Pipeline calls input.keys() instead of separate scan_keys()
- Streams are self-contained: read bytes + provide keys
- Rename DriveSession → Drive across entire codebase
- find_drives() returns Vec<Drive>, find_drive() returns Option<Drive>
- resolve_device() now pub(crate) — internal only
- StreamUrl is now a typed enum (Disc, Mkv, M2ts, Iso, Network, Stdio, Null)
with scheme() and path_str() accessors, replacing struct of Strings
- Add lock_tray() / unlock_tray() for safe disc access during rips
- Improve reset() with eject cycle that clears LibreDrive stuck state
- Add Send bounds to ScsiTransport and PlatformDriver traits
- DiscOptions uses PathBuf instead of String for device/keydb paths
- Update doc example to use new Drive API
- DriveStatus enum: TrayOpen, NoDisc, DiscPresent, NotReady, Unknown
- drive_status(): GET EVENT STATUS NOTIFICATION with TUR fallback
- reset(): PREVENT ALLOW → START STOP → init() escalation
- wait_ready(): tries reset on Illegal Request, falls back to drive_status
- Remaining: standard READ(10) still fails in LibreDrive stuck state
- wait_ready: detects MMkv vendor probe when TUR returns Illegal Request
- scan: capacity fallback to 0 when READ CAPACITY fails
- Still needs: re-init to restore standard SCSI commands, or use raw reads for UDF
- Doc comments on DriveSession, find_drives, all Error variants, Result type
- 24 format! strings inlined (clippy pedantic)
- 25 long hex literals with separators (0xFFFFFFFF → 0xFFFF_FFFF)
- README install example updated to 0.8
- SectorReader trait decouples disc scanning from SCSI
- Disc::scan_image() for ISO and any sector source
- resolve_encryption() handles AACS 1.0/2.0/none in one path
- IsoStream: full UDF/MPLS/CLPI/labels pipeline from ISO files
- StdioStream: stdin/stdout pipe
- Strict scheme:// URL format with validation
- Labels module refactored to SectorReader
- 7 stream types total
- DriveSession::open() no longer requires profile match — works on any optical drive
- init()/probe_disc() return error gracefully for unknown drives
- find_drives() returns all optical drives (PDT 0x05), not just profile-matched
- has_profile() check for callers
- AACS 2.0: handshake wired into resolve_aacs() — real VID + read_data_key
- DriveId: added raw_gc_010c field for GET_CONFIG 010C response bytes
- Split AacsError { detail } into 13 specific error variants (AacsCertShort,
AacsAgidAlloc, AacsCertRejected, etc.) with unique error codes E7001-E7012
- Split DiscError { detail } into 7 specific variants (DiscRead, MplsParse,
ClpiParse, UdfNotFound, DiscNoTitles, DiscTitleRange, DiscNoExtents)
- Add WriteError (E5001), KeydbLoad (E8005), MuxLookahead (E9000), MuxWrite (E9001)
- Add OpenDisc API for single-call open+scan+rip workflow
- Remove all English text from error Display impl (code-only output)
- Normalize doc comments to use -- instead of em dash for ASCII consistency
Firmware payload was extracted 12 bytes too early — included handler
table pointers instead of actual microcode. Fixed by scanning for
function table position (VM address pattern) instead of magic offsets.
All 206 profiles regenerated. Cold boot firmware upload should now work.
Two MPLS parsing issues:
1. PGS subtitle (0x90) appearing in audio stream slot: language was read
from audio offset sa[2..5] instead of PG offset sa[1..4], causing
truncation ("ng " instead of "eng"). Fixed by detecting PGS coding
type in audio slots and using correct offset.
2. Secondary PG (PiP subtitle) entries were not consumed, causing
position tracking to drift. Added missing n_pip_pg loop.
3. Added stream_type 5/6/7 attribute parsing for secondary audio/video.
Platform trait is no longer publicly exported. External code uses
DriveSession only — cannot call unlock, load_firmware, calibrate directly.
- Platform trait: pub(crate) with 3 methods only
- All handlers are private methods on Mt1959
- DriveStatus moved to mt1959 internal struct
- init() has guard: no re-init if already ready
- set_read_speed() has guard: no-op if not calibrated
- Removed open_unlocked() — open() is the only entry
- Removed Platform and DriveStatus from public exports
Prevents: out-of-sequence SCSI commands, double-init, wrong firmware writes.
- Removed open_unlocked() — open() is the only entry
- Removed redundant init() call from open_title()
- init() called once in open(), handles everything
- Each function does one thing: open→init→scan→read
status() (sub_cmd 0x13) returns ILLEGAL REQUEST on some drives.
Was retrying 6× with 30s timeouts = 180s hang during init().
Steps 1-10 all pass on hardware:
unlock: OK, load_firmware: OK, calibrate: OK,
register_a: OK, register_b: OK
Only status fails — not needed for reads.
When MPLS STN table parsing drifts (disc-specific alignment issue),
a PGS subtitle entry (coding_type 0x90/0x91) can appear in the audio
stream section. Previously this showed as garbled "ng PGS 5.1" audio.
Fix: guard in stream builder checks if audio-typed streams have
subtitle codecs and reclassifies them as subtitles.
Also: unknown stream types now filtered out (filter_map) instead of
creating fake Video entries that showed as "?" in output.
Tested on V for Vendetta BD — "ng PGS 5.1" gone, clean output.
Generated by: profile generator --profiles sdf0.bin keys.json --drive-db drive_profiles.json
206 profiles (66 A + 140 B), all with identity from brute-force dictionary.
No manual merging. One tool, three inputs, complete output.
A (): single WRITE_BUFFER → verify 0x45 → unlock×2
B (): WRITE handshake → READ 0x3000 → WRITE 16B → verify → unlock×5
9/10 handlers are identical A/B. Only load_firmware has different logic.
Both paths end with do_unlock() — firmware upload is a prerequisite for
unlock, not a substitute. init() tries unlock first, falls back to
load_firmware only on failure (cold boot).
Every handler traced instruction-by-instruction from operation: do_unlock with configurable response size
operation: WRITE_BUFFER + verify buf=0x45 + unlock×2
operation: do_unlock → validate → send pre-built CDB → [4:20]
operation: same with CDB B
operation: init → scan 0x0000-0x5800 → build table → triple speed
operation: ↔x86 VM only (host_write 16B), no SCSI
operation: do_unlock → validate → probe 0x13 → check sig → features
operation: 3 paths by param count (1/5/9), dynamic READ_BUFFER
operation: search 64-entry table → position probe →
set_cd_speed_max → custom SET_CD_SPEED with matched value
operation: ↔x86 VM only (host_read 8B), no SCSI
init() matches x86 dispatch exactly:
Phase 1: unlock → [load_fw] × 6
Phase 2: calibrate × 6
Phase 3: probe (drive info)
Phase 4: register A + B × 5
Phase 5: status × 6
Handlers 5/9 are VM communication (no SCSI equivalent in Rust).
All other handlers send real SCSI commands.
DriveProfile now has every field traced from firmware:
- drive_signature, unlock_init_value, unlock_response_size_minus_init
- ld_microcode (base64, ~1888B firmware payload)
- hardware_register_a_cdb, hardware_register_b_cdb (10B pre-built CDBs)
- drive_nominal_speed_cdb (12B calibration speed)
- speed_zone_table (28B), speed_calc_table (25B)
drive.rs simplified:
- open() calls init() instead of unlock()
- init() is the ONLY entry point — handles full dispatch sequence internally
- Removed read_config, read_register, maintain_speed, read_sectors from public API
- Added set_read_speed() for per-zone speed during content reads
- disc.rs updated to call init() instead of unlock()
Compiles clean, all tests pass.
MattJackson