Backfills W1, which was skipped when it merged, and adds W2's two verbs.
Notes the picker deviation: GetUserFileName both directions, not the
spec'd Win32/SWELL split.
Union of two parallel tracks. Both action rows, both menu rows, both link
edges survive; the two package CLAUDE.md files now describe the post-merge
reality rather than either side's pre-merge scope.
Delegates the refuse-gate to ledgerDegraded(), lifts its console message into a
pure testable fold, fixes stale doc line citations and an inaccurate outcome-enum
comment, and splits the rename counter into collision-vs-sanitize.
Clamps insertSuffix's underflow, floors uniqueEntryName's validity guard, suppresses
the redundant overwrite confirm via a picker out-param, adds a PayloadBuffer
high-water mark, and corrects stale CLAUDE.md/CMake claims.
Four collisions answered explicitly: ids reminted, names never overwritten,
content deduped before the write, bank name auto-suffixed. Degraded ledger
refuses before the picker.
Pure planner classifies missing/unreadable/unrepresentable and repairs transport
names; the verb digests, streams and commits atomically over a const session.
Notes the export verb's overwrite-consent obligation post-append, the append's
extension-divergence behavior, and readFilePayload's 4GiB blind spot; fixes a
stale u8string() reference and marks the 4GiB guard as accepted-unexercised.
Documents why parseSlots skips repeat-key rejection, corrects two drifted doc
lines (naming-rule count, Malformed-after-header header validity), and records
two forward obligations for import_plan in CLAUDE.md.
Cap readRange's allocation and reject size_t overflow instead of truncating;
re-append .rsbank when the export picker omits it; add cafe coverage for
writeFileExclusive and writeLandedFile; loop write() on EINTR.
Enum comment claimed package-id lookup that no persisted field supports; reworded
to the real distinction. Trimmed CLAUDE.md-duplicated test comments and the header.
Both pickers now ride GetUserFileName (mode 0/1); the "no save picker" premise was false.
Landing uses O_EXCL so the create is the existence check, not a TOCTOU pair.
An additively-tagged newer package that fails to parse now reports TooNew (with
writer semver) instead of unactionable Malformed. Format layer also refuses encoding
a zero-length entry, honoring the shell's appendPayload contract; both test-covered.
Reject NUL/control bytes and Windows-hostile names in entry names, relax
the over-broad ".." substring ban to component-only, close the
trailing-garbage gap on empty manifests, and relocate the package
CMake subdirectory to its ladder home.
Two-integer ladder (formatVersion/minReaderVersion), bare-name-only entries
validated on encode and decode, prefix decode that proves exact file size
without ever reading a payload.
Replace the self-comparing render-window loop with a genuinely discriminating
floor-vs-exact check; correct two stale claims; mark the Auto/Manual floor-parity
premise as unverified; drop the STARTPOS/ENDPOS comment's circular justification.
The millisecond floor lives in the custom-bounds field, not the engine, so
RENDER_BOUNDSFLAG=2 is now the only bounds mode: the two-position type, the
console verdict and the STARTPOS/ENDPOS drift probe all go. capture.cpp 697 -> 622.
Grid-ness of an edge was a proxy for "no floor could explain this count," not
the test itself — equal remainders on both edges cancel under a full floor. Now
checks all three floored models directly and corrects the SHORT/LONG
floor-signature docs.
Verdict can no longer print a false EXACT on an on-grid end, no longer names a
bounds channel a content-derived render never consulted, and the grid-align doc
premise is corrected without implementing it.
Corrects a false comment example, fixes two tests that couldn't detect their
own regressions, adds two more read-back checkpoints around Main_OnCommand so
a drift report self-locates, guards a spurious zero-vs-zero coincidence match,
and softens two sentences that overclaimed cause or defect.
No truncation exists on our side of that boundary, so the read-back is the
only evidence available for whether REAPER kept the window — and it fires on
every tail mode, where only None was ever judged.
REAPER's GetUserFileName serves both pickers — the no-save-picker finding was a
regex miss. A new enum value is structural, not additive: BankModel rejects an
out-of-range SourceMode/Tier rather than degrading.
The dedup path claimed it directly, so a key could be answered Ok pointing at an entry a failed persist never wrote. Write-back is now guarded and three-valued, and landing one request split out of the scan pass.
SetProjExtState's return covers the whole extname, so it never saw one key. The persist verdict now reaches the report, and a throw mid-write no longer claims the landing left nothing behind.
Splits Ignore into unreadable vs not-a-request and counts every verdict; the report prints only when the pass answered nobody, so its absence proves the action never ran.
Batch captures now name the retained-render folder once instead of nothing; Auto/Manual tail modes refuse a 0-byte render like None does; VERIFICATION.md steps 1-3 no longer invite a false conclusion; docs/comments no longer overclaim.
daniel