Update the --tui/path-display plan after review and live confirmation
- Corrected warning text: the empty-checklist message must not claim there are
no Missing/Broken/Mismatch candidates, since they can all have been excluded
as alias collisions.
- --showspecialfolder gains the two rules review found missing: extended-length
(\?\) paths are matched in their non-extended form, and diagnostics that
embed a path are abbreviated too.
- Verification updated to 491/493 and to record the reporting user's live
confirmation of the checklist, which closes the one item the implementation
session could not verify for lack of a real console.
- New "Review follow-up" section summarising what the review found.
5553fe2
Add the --tui reporting and path display plan (#179, #180)
New page (EN + JA) recording both post-M9 follow-ups: the fix --tui checklist
now listing Mismatch candidates as non-selectable informational rows, and the
new --showspecialfolder/-s option.
Also appends a post-implementation revision note to the M7 page (EN + JA)
carrying the full LinkStatus table - which statuses appear in the checklist,
which are selectable, and what fix does to each - because "fix never repairs a
Mismatch" is the part that is easy to forget, and #179 is what it looks like
when it is forgotten.
Indexed from Home, Home_ja and plan/syncwingetlink.
7b540c0
docs: add categorized milestone table to Home and Japanese plan translations
Home.md/Home_ja.md now list all eight registered milestone plans
(M2-M9) grouped by area, with date, status, tracking issue link, and a
language cross-link, matching the format used in the
vsextensionforcodex wiki. Each plan page gains a _ja.md counterpart
(none existed before). Home also gains a curated Microsoft Learn
reference section grouped by milestone area, covering the Win32/COM/
MSVC surface these plans implement.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
1d4f2a3
add: M9 COM API documentation plan (issue #11)
9d5f0c3
update: correct M8 plan for M7 merge, ADR renumbering, and #106/#118 findings
96298c4
fix: correct M7 TUI plan defects found in pre-implementation review
- Split the doubly-reserved ADR-0026 into ADR-0026/0027/0028 (one per
sub-issue #58/#59/#60), freeing #113 to use ADR-0029.
- Correct the exit-code precedence claim: the merged CLI already orders
interrupted ahead of insufficient-permission, contradicting "existing
precedence is preserved". M7 now explicitly fixes CLI-wide precedence
to permission(2) > interrupted/failure(10) > success(0), gated by
remaining > 0 for the interrupted case.
- Add a "declined" outcome distinct from dry-run "planned" so a refused
interactive confirmation is never reported as a dry-run plan.
- Specify the exact stdin input-mode flag set needed for resize
detection and Ctrl+C-as-key handling (ENABLE_WINDOW_INPUT vs.
ENABLE_VIRTUAL_TERMINAL_INPUT, ENABLE_EXTENDED_FLAGS for quick-edit,
srWindow-derived viewport instead of dwSize).
- Document terminal session init/teardown ordering and
close/logoff/shutdown-event restoration, since C++ destructors do not
run on those.
- Clarify stdout VT stays owned by Console; the new terminal session
owns only stdin mode, alt-screen, and cursor visibility.
35bc9d9
[add]: register M7 interactive TUI plan
8834eb9
add: M8 plan page, and correct M3/M6 pages against the merged code
Adds plan/syncwingetlink/m8-quality-polish-and-release, the design for issue
#10, and indexes it. The page records the divergences measured against the
merged M6 code (PRs #107-#112) with an owning sub-issue for each, a release
honesty contract, and the delivery sequence for #113, #63, #61, #62, #106,
#64, and #65.
Corrects two existing pages that no longer matched the code:
- M6: status was still "Planned, reviewed" although #8 and all its
sub-issues are closed; the rules-hardening item carried a placeholder
instead of #105; and the predicted ADR list named two ADRs with the wrong
subjects, where docs/adr-phase-5.md actually records six (ADR-0020 through
ADR-0025). Also drops the claim that #40 owns test-rule's presentation,
which #56 delivered itself.
- M3: #40 was described as structurally blocked. It was closed as already
implemented by #56's runTestRule(), so the page says so instead.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
fdaec84
fix: correct exit-code claims about non-interactive confirmation refusal
f49c674
fix: correct M6 path-override validation scope after implementing ArgParser
1b2f522
add: M6 command-line interface plan page
1ce29ff
[add]: document M5 symlink repair design
946ccbf
docs(wiki): mark M4 (link state inspection) as completed
Issue #6 and sub-issues #44-48 are all closed, merged via PRs
#96-#100.
ab41eb6
docs(wiki): fix M4 ADR filename to adr-phase-3.md and clarify NT-path decoding
adr-phase-4.md would leave a permanent phase-2->phase-4 gap since ADR files
are numbered sequentially by file, not by milestone, and adr-phase-2.md
already holds the M3 decisions. Also clarify that the existing
fromExtendedLengthPath helper does not cover the NT-namespace (\??\) prefix
a symlink SubstituteName uses, so decoding it is new M4 code, not reuse.
7c8caf5
[add]: add M4 link state inspection plan
6bf696a
Fold in M3 plan review findings (blockers, malformed-rules-file decision, doc corrections)
cfa002e
[add]: register the M3 alias resolution plan
0627341
Add M2 package source selection plan
1cf9a14