@feigi/fleet-ctl 3.24.2 → 3.24.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,10 +1,10 @@
1
1
  #!/bin/sh
2
- # Did the controller's instrument set change under it? (#436)
2
+ # Did the controller's instrument set change under it?
3
3
  #
4
4
  # The controller reads the world through scripts and runbooks that live in the
5
5
  # MAIN checkout, and every member can write to that checkout — except through a
6
- # `write`/`edit`/`ast_edit` that member-write-guard.mjs refuses (ADR 0020); a
7
- # `bash` or `eval` write still lands, the backstop #2210 tracks. Worktree isolation
6
+ # `write`/`edit`/`ast_edit` that member-write-guard.mjs refuses; a
7
+ # `bash` or `eval` write still lands. Worktree isolation
8
8
  # and `./agent-test` port derivation protect members from each other; neither
9
9
  # protects the tree the controller measures from. Measured: a member edited two
10
10
  # files in the main checkout instead of its worktree while the CI monitor was
@@ -42,7 +42,7 @@
42
42
  # already knows can be stale). A list rots on the next commit that adds a
43
43
  # script; a directory does not.
44
44
  #
45
- # NOT THE WHOLE REPO, deliberately: since #1336 the plugin's payload is
45
+ # NOT THE WHOLE REPO, deliberately: the plugin's payload is
46
46
  # nested under `plugin/`, and the repo root also carries `docs/`, `.github/`
47
47
  # and `.out-of-scope/`. `docs/metrics/` is APPENDED BY THE CONTROLLER
48
48
  # MID-RUN, so a whole-repo set would fire the gate on the run's own
@@ -65,7 +65,7 @@
65
65
  # blob untouched; both exit 0 here. A mode change that costs the hash its READ
66
66
  # is caught instead: `chmod 000` on a tracked instrument refuses at exit 2
67
67
  # through `could not hash every tracked file` — as root that never fires,
68
- # since root reads a mode-000 file regardless (instruments.test.mjs skips
68
+ # since root reads a mode-000 file regardless (the suite skips
69
69
  # this case there for the same reason). So the exposure is the bits that
70
70
  # still permit reading, the exec bit above all — and that bit is not how a
71
71
  # fleet probe runs: fleet-run hands `.sh` to `sh` and `.mjs` to `node`, so a
@@ -74,20 +74,19 @@
74
74
  # modes in — digesting `git ls-files -s` alongside the contents, or a per-file
75
75
  # `[ -x "$f" ]` — is a behaviour change to the gate on a channel with no
76
76
  # silently wrong reading to its name, and whether a mode-only edit is worth
77
- # refusing on is a decision rather than a fix, so #1059 names the gap and
78
- # leaves the behaviour alone. instruments.test.mjs pins the accepted pair,
77
+ # refusing on is a decision rather than a fix, so this paragraph names the gap and
78
+ # leaves the behaviour alone. The suite pins the accepted pair,
79
79
  # so closing the gap later goes red there rather than leaving this
80
80
  # paragraph stale.
81
81
  #
82
82
  # REFS ARE DELIBERATELY NOT IN THE DIGEST, and the reason is not that they do
83
- # not matter. The corroborating evidence on #436 is a stray `fix/42-slug` branch
83
+ # not matter. The corroborating evidence is a stray `fix/42-slug` branch
84
84
  # left in the main checkout by something running this repo's own fixtures — a
85
85
  # ref write, which no hash of files can see. But `claim-ticket.sh` creates a
86
86
  # branch per ticket and `reap.sh` deletes them, in the ref store every worktree
87
87
  # shares, so a ref digest changes several times per pass as ORDINARY WORK. Per
88
88
  # gate that is noise, and noise is the one failure this check cannot afford.
89
- # Sweeping for unexpected refs is a drain-cadence question, not a gate one, and
90
- # it is not what #436's acceptance criteria ask for.
89
+ # Sweeping for unexpected refs is a drain-cadence question, not a gate one.
91
90
  #
92
91
  # COST: two `git ls-files` (the state home's own set, which decides whether a
93
92
  # recorded root may be followed at all, and the audited tree's set) and one
@@ -98,19 +97,19 @@
98
97
  set -eu
99
98
 
100
99
  # Directly below `set -eu`, not below a locale pin: this script has none and is
101
- # deliberately off locale-pin-prose.test.mjs's PINNED list. The five siblings
102
- # that DO carry a pin put this line under it instead, because that file's
103
- # PROLOGUE regex admits only comments, blanks and `set -[eux]+` above the pin.
100
+ # deliberately off the suite's list of locale-pinned scripts. The siblings
101
+ # that DO carry a pin put this line under it instead, because the suite admits
102
+ # only comments, blanks and `set -[eux]+` above the pin in those scripts.
104
103
  # Here the only constraint left is the real one: above the first git call.
105
104
  #
106
- # GIT_WORK_TREE is #1337's defect reached through the environment. The
107
- # contract that ticket established is that the audited tree is the WORKING
105
+ # GIT_WORK_TREE is the wrong-tree defect reached through the environment. The
106
+ # contract here is that the audited tree is the WORKING
108
107
  # DIRECTORY's checkout — and `git rev-parse --show-toplevel` answers with the
109
- # ambient work tree instead the moment one is set, `--repo` or not. Measured
110
- # (#1020): standing in a checkout whose instruments have been TAMPERED, with
108
+ # ambient work tree instead the moment one is set, `--repo` or not. Measured:
109
+ # standing in a checkout whose instruments have been TAMPERED, with
111
110
  # `GIT_WORK_TREE` naming a clean twin that carries its own pinned baseline,
112
111
  # this script exits 0 and prints the twin's digest. The gate passes. That is
113
- # the same wrong-tree write `--repo ""` produced before #1350 refused it, one
112
+ # the same wrong-tree write `--repo ""` produced before this script refused it, one
114
113
  # door further out, and a gate that certifies a tree nobody looked at is
115
114
  # worse than no gate.
116
115
  #
@@ -122,12 +121,12 @@ set -eu
122
121
  # pair is one hazard with one remedy, and because "inert today" is a
123
122
  # measurement of the current call set: a digest taken from `git show` or
124
123
  # `cat-file` rather than from disk would reintroduce the half nothing here
125
- # can see. ambient-git-vars-prose.test.mjs pins the line itself, which is
124
+ # can see. The suite pins the line itself, which is
126
125
  # what keeps that half from being quietly dropped.
127
126
  unset GIT_DIR GIT_WORK_TREE
128
127
 
129
128
  NAME=instruments
130
- # `printf '%s'`, never `echo` (#484): `echo` expands backslash escapes in its
129
+ # `printf '%s'`, never `echo`: `echo` expands backslash escapes in its
131
130
  # operand, and the messages below carry paths that git will happily hand us with
132
131
  # a backslash in them.
133
132
  die() { printf '%s: %s\n' "$NAME" "$1" >&2; exit 2; }
@@ -154,9 +153,9 @@ while [ $# -gt 0 ]; do
154
153
  [ $# -ge 2 ] || die "usage: instruments.sh [--pin] [--repo <path>] [--audit <path>]"
155
154
  repo=$2
156
155
  # An empty value falls through the `[ -n "$repo" ]` branch below and
157
- # silently re-derives from cwd — measured (#1350 review): `--repo ""`
156
+ # silently re-derives from cwd — measured: `--repo ""`
158
157
  # pinned the CALLER's cwd repo instead of refusing, exactly the
159
- # wrong-tree-write class this ticket exists to close. Refuse here,
158
+ # wrong-tree-write class this script exists to close. Refuse here,
160
159
  # before that check ever runs.
161
160
  [ -n "$repo" ] || die "--repo requires a non-empty path"
162
161
  shift 2
@@ -171,10 +170,10 @@ done
171
170
  #
172
171
  # `root` — the STATE HOME — is the WORKING DIRECTORY's checkout, never the
173
172
  # checkout this script happens to ship from. Under the install-only dev loop
174
- # (ADR 0003) this file runs out of a plugin cache — on a real install, under
173
+ # this file runs out of a plugin cache — on a real install, under
175
174
  # `~/.omp/plugins/cache/...` — and the OLD own-location contract (`git -C
176
175
  # "$(dirname "$0")" …`) resolved it to whatever git checkout happens to
177
- # CONTAIN that cache path. Measured (#1337), pre-cutover: on a real box that
176
+ # CONTAIN that cache path. Measured, pre-cutover: on a real box that
178
177
  # was the operator's unrelated personal dotfiles checkout, and `--pin` run
179
178
  # that way writes `.fleet/instruments.sha` into it. `--repo <path>` is the
180
179
  # explicit override for the one legitimate case that needs a home other than
@@ -209,7 +208,7 @@ base="$root/.fleet/instruments.sha"
209
208
  # Whether the state home tracks an instrument set OF ITS OWN decides if a
210
209
  # second tree may be named at all — and it is what closes the door a recorded
211
210
  # root would otherwise open: measured, a member who can write `.fleet/`
212
- # (member-write-guard does not stop `bash`, #2210) can keep the honest
211
+ # (member-write-guard does not stop `bash`) can keep the honest
213
212
  # run-start digest on line 1 and move `root=` to a pristine clone of the
214
213
  # workspace whose bytes match it — the old script hashed the tampered tree and
215
214
  # refused; a check that trusted the pointer would certify the clone and pass.
@@ -228,7 +227,8 @@ if [ -n "$audit" ]; then
228
227
  target=$(git -C "$audit" rev-parse --show-toplevel) \
229
228
  || die "$audit is not inside a git checkout — cannot identify the instrument set to pin"
230
229
  elif [ "$pin" = true ] && [ -z "$home" ]; then
231
- # #1058's refusals mirrored on this path: a baseline that EXISTS but cannot
230
+ # The check path's unreadable-baseline refusals, mirrored here: a baseline
231
+ # that EXISTS but cannot
232
232
  # be read names the tree this re-pin is supposed to keep certifying, and
233
233
  # overwriting it blind would discard the pointer with no record that it was
234
234
  # ever there.
@@ -260,21 +260,20 @@ if [ "$pin" != true ]; then
260
260
  # case "no baseline" and prescribing `--pin` would walk a controller from
261
261
  # "the check could not look" to "certified clean" in one step, discarding
262
262
  # evidence it never read — the exact anti-pattern run-team/SKILL.md forbids.
263
- # (#1058)
264
263
  basedir="$(dirname "$base")"
265
264
 
266
265
  # A directory that exists but cannot be searched (missing +x) hides
267
266
  # everything under it from stat(2) — `[ -e "$base" ]` below reads FALSE for
268
267
  # every file underneath, so without this check the run falls through to the
269
268
  # "no baseline" message one level up: the exact mislabel this refusal exists
270
- # to prevent, just moved from the file to its containing directory. (#1058)
269
+ # to prevent, just moved from the file to its containing directory.
271
270
  [ -d "$basedir" ] && [ ! -x "$basedir" ] \
272
271
  && die "$basedir exists but is unreadable — fix its permissions; do NOT --pin over it, --pin overwrites rather than compares"
273
272
 
274
273
  # `-L` catches a dangling symlink: the link entry is present but its target
275
274
  # is gone, so `-e` (which dereferences) reads FALSE and the run would
276
275
  # otherwise fall through to the same "no baseline" message for a baseline
277
- # that is very much present, just broken. (#1058)
276
+ # that is very much present, just broken.
278
277
  { [ -e "$base" ] || [ -L "$base" ]; } && [ ! -r "$base" ] \
279
278
  && die "$base exists but is unreadable — fix its permissions; do NOT --pin over it, --pin overwrites rather than compares"
280
279
  [ -r "$base" ] || die "no baseline at $base — run instruments.sh --pin once at run start"
@@ -345,7 +344,7 @@ git -C "$target" ls-files -z -- $set > "$files" \
345
344
  # run at all leaves the second one hashing empty input and exiting 0, and the
346
345
  # empty-input digest then compares unequal and reads as CHANGED. That is exit 1
347
346
  # — a verdict about the tree — off a failure to look. reap.sh shipped this exact
348
- # bug against `git cherry` (#264) and it deleted branches.
347
+ # bug against `git cherry` and it deleted branches.
349
348
  #
350
349
  # `xargs` exits 123 when `shasum` failed on any file, which is how a tracked
351
350
  # file deleted from the worktree arrives: as exit 2 naming it, not as a digest
@@ -382,7 +381,7 @@ fi
382
381
 
383
382
  # Cold path only, so it costs the ordinary run nothing. The controller has to
384
383
  # report WHAT changed, and neither digest says. `git status` names the
385
- # uncommitted half, which is the shape the near-miss on #436 actually took; a
384
+ # uncommitted half, which is the shape this header's measured near-miss took; a
386
385
  # checked-out branch that moved shows as nothing here and the HEAD line is what
387
386
  # names it. Both are read from the AUDITED tree, which is the one the verdict
388
387
  # is about — not the state home the baseline was found in.
package/scripts/json.sh CHANGED
@@ -30,14 +30,14 @@
30
30
  # reads and has a SYNTAX error aborts before either guard on /bin/sh, bash 3.2
31
31
  # and `bash --posix`, while bash 5.3 reaches the `|| die` with status 2. The
32
32
  # suite does NOT cover that case: measured, a syntax error appended at EOF
33
- # leaves bash 3.2 with every function it had already parsed and json.test.mjs
33
+ # leaves bash 3.2 with every function it had already parsed and the suite
34
34
  # stays green. CI's `shellcheck -x -S warning` over `git ls-files '*.sh'` is
35
35
  # what catches it (SC1072/SC1073).
36
36
  #
37
37
  # `LC_ALL=C` is a per-command prefix on every tool below, never an export. Three
38
38
  # of the callers (verify-sha.sh, prove-merge.sh, claim-ticket.sh) do not pin the
39
39
  # locale, and exporting from a sourced lib would silently re-locale every OTHER
40
- # tool in them. The prefix is a no-op in the five that do pin it (#582), and
40
+ # tool in them. The prefix is a no-op in the five that do pin it, and
41
41
  # closes the same hazard in the three that do not. It matters here because BSD
42
42
  # `tr` exits 1 on a byte that is not valid UTF-8 under a UTF-8 locale, and
43
43
  # `sed` emits nothing at all.
@@ -52,7 +52,7 @@
52
52
  # zero-width word BOUNDARY to GNU sed and a literal `b` to BSD sed, so the rule
53
53
  # would insert `\b` at every word edge on one and mangle every letter `b` on the
54
54
  # other (measured, GNU sed 4.9 and macOS sed). \177 (DEL) is not a C0 byte and
55
- # JSON permits it unescaped, so it is left alone (#146). Every remaining byte
55
+ # JSON permits it unescaped, so it is left alone. Every remaining byte
56
56
  # below \040 has no short form, \013 (VT) included — RFC 8259 lists exactly the
57
57
  # five above and `\v` is not among them; `tr` turns it into a space, and
58
58
  # `jrewritten` is how a caller finds out that happened, since a replaced value
@@ -70,11 +70,11 @@
70
70
  # POSIX leaves this undefined and GNU sed's answer differs — so plain `N;$!ba`
71
71
  # prints nothing at all for a single-line value.
72
72
  #
73
- # EVERY FALLIBLE STAGE'S STATUS IS READ HERE, and that is the #119 fix rather
73
+ # EVERY FALLIBLE STAGE'S STATUS IS READ HERE, and that is a fix rather
74
74
  # than a style choice. A pipeline's status is its LAST stage's, so the
75
75
  # original `printf | sed | tr` reported only `tr` — forcing `sed` to fail left
76
76
  # jstr exiting 0 with an empty value and every caller's `|| die` unfired, while
77
- # forcing `tr` worked (measured on PR #425). POSIX sh has no `pipefail` and no
77
+ # forcing `tr` worked (measured). POSIX sh has no `pipefail` and no
78
78
  # `PIPESTATUS`, so each stage that can fail is captured and its status read.
79
79
  # The consequence, and the rule a later edit is checked against: a pipeline may
80
80
  # still END a function, but only where its LAST stage is the fallible one — `tr`
@@ -89,7 +89,7 @@
89
89
  # indistinguishable one and `$( )` takes it off — `jstr` of `a<LF>` and of `a`
90
90
  # are byte-identical (measured). No live caller can reach it: every argument
91
91
  # reaching jstr today is a script literal or a `$( )`/`awk`/`read` capture that
92
- # has already lost its own trailing newline. #119 leaves it there rather than
92
+ # has already lost its own trailing newline. This file leaves it there rather than
93
93
  # adding a sentinel no caller would exercise. jarr is the exception and is
94
94
  # handled in its own note.
95
95
  #
@@ -104,13 +104,13 @@
104
104
  # sequence starts or continues with — sailed through untouched and landed raw
105
105
  # in the payload: `jq` then silently substitutes U+FFFD on decode (changing
106
106
  # what the caller reads back) and a strict parser (`python3 json.load`)
107
- # refuses the payload outright (#613). `jstr` and `jarr` now run the value
107
+ # refuses the payload outright. `jstr` and `jarr` now run the value
108
108
  # through Python's own UTF-8 decoder with `errors="replace"` first — the same
109
109
  # U+FFFD a lenient consumer already substitutes, made explicit and JSON-legal
110
110
  # at the source instead of implicit and consumer-dependent — and `jrewritten`
111
111
  # reports it as a rewrite like any other replaced byte.
112
112
 
113
- # Python's own UTF-8 decoder with errors="replace" (#613), shared script text
113
+ # Python's own UTF-8 decoder with errors="replace", shared script text
114
114
  # rather than a function: a function called on the right side of a pipe runs
115
115
  # in its own subshell, and a `json_u8=` assigned inside it would not survive
116
116
  # back to the caller — every call site below runs this as
@@ -123,7 +123,7 @@ sys.stdout.buffer.write(sys.stdin.buffer.read().decode("utf-8", "replace").encod
123
123
  # Escape $1 into a JSON string BODY — no surrounding quotes, the caller adds
124
124
  # those. Exit 0 with the escaped value, non-zero if any stage failed.
125
125
  jstr() {
126
- # Empty in, empty out, no fork at all (#120). This is what lets a PATH-wide
126
+ # Empty in, empty out, no fork at all. This is what lets a PATH-wide
127
127
  # sed/tr outage — the failure inflight.sh measures at its own top — leave a
128
128
  # field that legitimately found nothing untouched: that field never calls the
129
129
  # broken tool, so it cannot observe its failure.
@@ -179,7 +179,7 @@ jrewritten() {
179
179
  # short-circuit above existed).
180
180
  json_ascii=$(printf '%s' "$1" | LC_ALL=C tr -d '\200-\377') || return 1
181
181
  [ "$json_ascii" = "$json_orig" ] && { printf false; return 0; }
182
- # jstr also repairs a byte that is not valid UTF-8 (#613) — a second,
182
+ # jstr also repairs a byte that is not valid UTF-8 — a second,
183
183
  # independent replacement alongside the C0 scrub above, so a second,
184
184
  # independent check: this value must decode as strict UTF-8 unchanged, or
185
185
  # jstr rewrote it too.
@@ -194,15 +194,14 @@ jrewritten() {
194
194
  # boundary between two elements by the time a value reaches per-line stdin,
195
195
  # which is why no-undo-audit.sh refuses a conflicting path holding one before it
196
196
  # ever calls this. Escaping a byte this function structurally never receives
197
- # would be dead code standing in for a restructure nobody has needed; #89 owns
198
- # that class.
197
+ # would be dead code standing in for a restructure nobody has needed.
199
198
  #
200
199
  # Empty stdin gives empty stdout at exit 0 — measured as the original
201
200
  # pipeline's behaviour, and the `[ -n ]` guard is what preserves it: without it
202
201
  # the `printf '%s\n'` re-emit below would turn "no elements" into one empty
203
202
  # line and `paste` would answer `""`, inventing an element out of nothing.
204
203
  jarr() {
205
- # Same repair as jstr, same reason (#613), run once over the whole
204
+ # Same repair as jstr, same reason, run once over the whole
206
205
  # (possibly multi-line) input rather than per line: \n is ASCII and never a
207
206
  # byte of a multi-byte UTF-8 sequence, so decoding the joined blob in one
208
207
  # pass cannot manufacture or hide a line boundary.
package/scripts/net.sh CHANGED
@@ -1,5 +1,5 @@
1
1
  # shellcheck shell=sh
2
- # The fleet's bounded, prompt-suppressed git transport (#92, #346, #347).
2
+ # The fleet's bounded, prompt-suppressed git transport.
3
3
  # Sourced, never executed — no shebang, and the `shell=sh` directive above is
4
4
  # what tells shellcheck what to check it as.
5
5
  #
@@ -16,19 +16,19 @@
16
16
  # builtin, so failing to open its operand aborts a non-interactive shell before
17
17
  # any `||` on the line can run.
18
18
  #
19
- # WHY A LIB. #346 landed this mechanism inside inflight.sh's probe 2, for the
19
+ # WHY A LIB. This mechanism first landed inside inflight.sh's probe 2, for the
20
20
  # one call that probe makes. Every other unattended `git` network call in the
21
21
  # fleet — release-ticket.sh's pushed-branch lookup, and the fetches in
22
22
  # prove-merge.sh, reap.sh and verify-sha.sh — still ran raw, so each of them
23
23
  # could prompt on a credential or a host key, or stall on a transport that
24
- # connects and then goes quiet, holding a fleet slot indefinitely (#347). The
24
+ # connects and then goes quiet, holding a fleet slot indefinitely. The
25
25
  # alternative considered was copying the transport flags to each call site; it
26
26
  # is what this file exists to avoid, and it would not have bounded the connect
27
27
  # phase in any case, since git exposes no knob for it. So the mechanism moved
28
28
  # here whole and probe 2 became its first caller rather than its owner.
29
29
  #
30
- # `gh` calls are deliberately NOT routed through this. Measured during the
31
- # PR #332 review, `gh issue view` against a silent listener failed on its own at
30
+ # `gh` calls are deliberately NOT routed through this. Measured: `gh issue
31
+ # view` against a silent listener failed on its own at
32
32
  # 10.0s with `net/http: TLS handshake timeout` — `gh` is already bounded, and
33
33
  # wrapping it would buy a second bound over the first.
34
34
  #
@@ -41,7 +41,7 @@
41
41
  # The override exists so the tests can buy a short budget instead of paying the
42
42
  # default one per case. It can only ever SHORTEN: a knob that could lengthen it
43
43
  # would be one more way for configuration to remove the bound, which is the
44
- # defect the ssh half of #346 reports, and reproducing it here to be convenient
44
+ # defect this lib exists to close, and reproducing it here to be convenient
45
45
  # would be its own bug. A value that is not a positive integer is not an error
46
46
  # and not a bound either — the default stands.
47
47
  #
@@ -100,11 +100,12 @@ net_fetch_budget() {
100
100
  # reparents its children and the links this walk follows are gone by then.
101
101
  #
102
102
  # Ceiling: with no `ps` to read, this falls back to the named process alone and
103
- # a helper can survive it. That is the pre-#346 behaviour for that one case, not
104
- # a new failure, and it is preferred over signalling a set derived from nothing.
103
+ # a helper can survive it. That is the old named-process-only kill for that
104
+ # one case, not a new failure, and it is preferred over signalling a set
105
+ # derived from nothing.
105
106
  # But it is not silent: the fallback records itself in $net_wdfile so the reason
106
107
  # printed downstream can say the bound may not have held. A degraded kill that
107
- # reads exactly like a clean one is the defect #346 asks this script not to
108
+ # reads exactly like a clean one is a defect this script must not
108
109
  # have — measured, `ps` shimmed to exit 127 produced byte-identical stderr to
109
110
  # the healthy run while a `git remote-https` survived and held the caller for
110
111
  # its whole cap. `${net_wdfile:-/dev/null}` because net_kill_tree also runs
@@ -126,7 +127,7 @@ net_fetch_budget() {
126
127
  # parents first, which is exactly why every end-to-end case here stays green on
127
128
  # that mutant.
128
129
  #
129
- # What holds it is the descending-pid chain in net.test.mjs's net_kill_tree
130
+ # What holds it is the descending-pid chain in the suite's net_kill_tree
130
131
  # case — each child's pid BELOW its parent's, the shape pid wraparound produces.
131
132
  # Feeding the rows in a different order does not hold it: the hash order is not
132
133
  # the input order, and measured, the same key set fed parents-first and
@@ -234,7 +235,7 @@ net_stalled() {
234
235
  # above: it gates the banner exchange, not only the TCP handshake, so a peer
235
236
  # that accepts the connection and then never speaks is cut off at
236
237
  # ConnectTimeout. Measured against that exact case (OpenSSH_10.2p1, the
237
- # accept-then-silent listener the test at inflight.test.mjs uses) — both
238
+ # suite's accept-then-silent listener) — both
238
239
  # options set: "Connection timed out during banner exchange" at 10.0s;
239
240
  # ConnectTimeout alone, ServerAlive dropped: 10.0s, identical; ServerAlive
240
241
  # alone, ConnectTimeout dropped: still running at 30s, killed from outside.
@@ -269,9 +270,9 @@ net_stalled() {
269
270
  # asked for, and terminates only where that value does. Measured on the same
270
271
  # listener: a user ConnectTimeout of 0 is accepted, wins by the same rule and
271
272
  # left the call still connecting at 40s, where the 10s set here cut at 10.2s —
272
- # unbounded, through the user's own config, which is the #92 hang again. The
273
+ # unbounded, through the user's own config, which is the unattended hang again. The
273
274
  # watchdog below is what now bounds that case, since it bounds the call rather
274
- # than the transport and so does not depend on any value ssh resolved (#346).
275
+ # than the transport and so does not depend on any value ssh resolved.
275
276
  # Ordering these first would bound the call at its own value instead,
276
277
  # at the cost of silently overriding a deliberate proxy or timeout config: a
277
278
  # real regression traded for a hypothetical one, so it is not done.
@@ -283,7 +284,7 @@ net_stalled() {
283
284
  # the BatchMode=yes set here alone, an unknown host key fails at once ("Host key
284
285
  # verification failed"); with a user's BatchMode=no ahead of it and a terminal
285
286
  # reachable, ssh sat on "Are you sure you want to continue connecting" until
286
- # killed — the unattended hang #92 exists to stop. Reachable is the operative
287
+ # killed — the unattended hang this lib exists to stop. Reachable is the operative
287
288
  # word: with no terminal available it aborts rather than waiting. Honoured
288
289
  # anyway, because it is the user's explicit setting; this records what that
289
290
  # costs rather than warning about a choice they made on purpose. The watchdog
@@ -291,7 +292,7 @@ net_stalled() {
291
292
  # the prompt is ssh's, so no git-side variable reaches it, and the setting that
292
293
  # opens it is the user's own, so overriding it is not on offer either. No test
293
294
  # covers this one — the exposure needs a reachable terminal, and the harness
294
- # runs without one, where ssh aborts at once instead of asking (#346).
295
+ # runs without one, where ssh aborts at once instead of asking.
295
296
  #
296
297
  # http: lowSpeedLimit/lowSpeedTime is git's (curl's) own bound for a transfer
297
298
  # that goes quiet — abort if it sits under 1000 bytes/s for 10s.
@@ -308,7 +309,7 @@ net_stalled() {
308
309
  # script does not set and cannot shorten.
309
310
  #
310
311
  # Hence the watchdog below, and hence these knobs are no longer the bound. They
311
- # stay anyway, and that is a decision rather than an oversight (#346): each one
312
+ # stay anyway, and that is a decision rather than an oversight: each one
312
313
  # fails earlier than the watchdog and in git's own words, which is the more
313
314
  # useful thing to read on a terminal, and dropping them would make every ssh
314
315
  # stall wait out the full budget where ConnectTimeout ends it in a fraction of
@@ -320,9 +321,10 @@ net_stalled() {
320
321
  # call — its own `sleep` is a child, and an orphaned sleeper does not sit
321
322
  # harmlessly: it wakes at the end of its budget and fires net_kill_tree at a pid
322
323
  # this shell no longer owns, which after a budget's worth of pid churn can be an
323
- # unrelated process — and, since #346, an unrelated SUBTREE. It cannot leak onto
324
- # the caller's stderr whatever else it does, because the `>/dev/null 2>&1` below
325
- # is on the sleeper itself and both its descriptors are already closed.
324
+ # unrelated process — and, because the kill walks every descendant, that
325
+ # process's whole SUBTREE. It cannot leak onto the caller's stderr whatever else
326
+ # it does, because the `>/dev/null 2>&1` below is on the sleeper itself and both
327
+ # its descriptors are already closed.
326
328
  #
327
329
  # `wait` is captured through an explicit `|| net_status=$?`, and `set -e` is
328
330
  # why: a bare `wait` on a killed child aborts at that line, which is after the
@@ -352,7 +354,7 @@ net_git() {
352
354
  # sends TERM then KILL with no pause, so whichever one reaps the job first is
353
355
  # a race, not a property either shell owns. What the shell DOES fix is the
354
356
  # FORM — bash-as-sh names the job (the long form); dash's notice is bare, just
355
- # `<signal>: <n>`. Measured for #1042 by replaying net_git's own mechanism
357
+ # `<signal>: <n>`. Measured by replaying net_git's own mechanism
356
358
  # against a plain `sleep`: on this machine, `/bin/sh` (bash 3.2.57) and
357
359
  # `/bin/dash` both gave TERM, every run — a quiet-system sample, not a signal
358
360
  # either shell is owed, which is why a caller has to accept either.
@@ -379,7 +381,7 @@ net_git() {
379
381
  # worktree directory, HEAD, or anything else in the entry — is pruned on
380
382
  # sight: the 3-month floor bounds staleness git can measure from the
381
383
  # entry's own files, and a missing pointer gives it nothing to measure,
382
- # so the floor does not apply. Measured (#2078): a fixture with exactly
384
+ # so the floor does not apply. Measured: a fixture with exactly
383
385
  # that fault, fetched through this function unmodified, loses the
384
386
  # registry entry between the fetch returning and the caller's own next
385
387
  # read of it — a worktree whose fault is seconds old, not months.
@@ -33,7 +33,7 @@
33
33
  # parent that looks right.
34
34
  #
35
35
  # The payload splits the checks into three kinds, because a flat object could not
36
- # say which of its fields were load-bearing for the verdict it sits next to (#18):
36
+ # say which of its fields were load-bearing for the verdict it sits next to:
37
37
  #
38
38
  # preconditions — every `die` above the verdict. They exit 2 with NO payload, so
39
39
  # they never appear as a false field beside `proved`. `merge ∈
@@ -55,7 +55,7 @@ set -eu
55
55
  NAME=prove-merge
56
56
  die() { printf '%s: %s\n' "$NAME" "$1" >&2; exit 2; }
57
57
 
58
- # The escaping helpers (#119). json.sh's header holds the sourcing contract and
58
+ # The escaping helpers. json.sh's header holds the sourcing contract and
59
59
  # the measurements behind it. Exit 1 out of THIS script means "the proof is a
60
60
  # no", which the merge bot reads as a reason to refuse a merge, so a missing
61
61
  # library must not be able to say it — which is why `[ -r ]` has to fire before
@@ -66,7 +66,7 @@ json_lib="$(dirname "$0")/json.sh"
66
66
  # shellcheck source=json.sh
67
67
  . "$json_lib" || die "$json_lib failed to load"
68
68
 
69
- # The bounded, prompt-suppressed git transport (#92, #346, #347). The fetch
69
+ # The bounded, prompt-suppressed git transport. The fetch
70
70
  # below is unattended: with no bound it can prompt for a credential or a host
71
71
  # key, or stall on a transport that connects and then goes quiet, and either
72
72
  # holds a fleet slot until something outside kills it. net.sh's header holds the
@@ -116,7 +116,7 @@ for obj in "$pre" "$post" "$merge"; do
116
116
  git cat-file -e "${obj}^{commit}" || die "cannot resolve $obj to a commit in this repository"
117
117
  # -e above peels through a tag to reach the commit underneath, so an
118
118
  # annotated tag clears that line even though $obj itself is not a commit —
119
- # only the object at the far end of the peel is (#585). $pre and $post are
119
+ # only the object at the far end of the peel is. $pre and $post are
120
120
  # then resolved UNPEELED by plain `git rev-parse` below, so a tag there hands
121
121
  # back the tag object's own sha rather than the commit's, and the identity
122
122
  # test later in this script compares that against a commit sha and can never
@@ -139,7 +139,7 @@ echo "\$ git merge-base --is-ancestor $merge $base" >&2
139
139
  # Assigned first, not tested inline: `die` inside `$(...)` exits the subshell,
140
140
  # and `[ ]` discards that status, so `set -e` never fires and a probe that could
141
141
  # not answer falls through to the disproof below — a verdict never established.
142
- # An assignment is a simple command, so its status is the substitution's (#267).
142
+ # An assignment is a simple command, so its status is the substitution's.
143
143
  merge_anc=$(is_ancestor "$merge" "$base")
144
144
  [ "$merge_anc" = true ] || die "$merge is not reachable from $base — that merge did not land"
145
145
 
@@ -213,8 +213,8 @@ echo "$NAME: proved=$proved (path=$proof_path)" >&2
213
213
  # come from `git rev-parse`, 40 hex characters and nothing else, and
214
214
  # `$proof_path` is this script's own `rebase`/`no-rebase` literal. They go
215
215
  # through `jstr` for uniformity against a later edit that changes where a field
216
- # comes from, the same reason inflight.sh wraps `$pr`, and it costs nothing
217
- # (#119). Assigned before the printf, never inline in its argument list — a
216
+ # comes from, the same reason inflight.sh wraps `$pr`, and it costs nothing.
217
+ # Assigned before the printf, never inline in its argument list — a
218
218
  # `$()` there sits outside this `|| die`, contributes an empty argument on
219
219
  # failure, and printf still exits 0 with a malformed payload.
220
220
  second_j=$(jstr "$second") && first_j=$(jstr "$first") && path_j=$(jstr "$proof_path") \