[zero-aska]

This is a follow-up to #44. #44 was merged before its independent review finished. That post-merge review said merge, but found that #44's own fix was incomplete:

• Newman and Ziff's p_c was relabelled as a measurement in the tests and in two READMEs. It was still called exact in the study's results table and in critical/table.mjs.
• Salas and Sokal's V4 = 1.1679229(47) is also a high-precision value rather than an exact one. It comes from exact CFT correlators integrated numerically, and its error is 1.6e-6 in U, which is negligible here.
• The site's turbulence entry still implied that viscosity always ends the cascade.

No number, tolerance or check changes.

Status

• 7410d00 made the first round of changes. Its review found one blocking item: the new table.mjs header named only p_c and V4 as not exact.
• 60add3c fixed that (reply). Its re-review found one new blocking item: the turbulence bullet's k^(−5/3) sentence had come to cover the solar wind and molecular clouds.
• 31d0f92 fixed that (reply). Its review said merge, but found four wording slips that 31d0f92 itself introduced.
• 1ad7386 fixes those (reply). Its independent check said merge, with no blocking findings. CI is green on Node 18, 20 and 22. The check's two non-blocking items were about this description and a reply, and both are corrected.

Changes, by commit

• 7410d00
  • critical/README.md:
    • The "What it reproduces" table heads its reference column "Exact or high-precision".
    • The short answer says the model "lands on or near the known numbers".
    • The opening paragraph names V4 beside Newman and Ziff's threshold.
  • critical/table.mjs:
    • p_c is printed as Newman & Ziff's measurement, and V4 as Salas & Sokal's high-precision value.
    • The pooled percolation means print reference for p_c and exact for 91/48.
  • tests/critical.mjs: the two Binder-cumulant checks print V4 as reference, and the header calls the cumulant high-precision.
  • index.html: the turbulence entry says what sets the smallest size depends on the medium: viscosity in air and water, the ions' own motion in the solar wind.
• 60add3c
  • critical/table.mjs:
    • The first output block is headed "Known values: exact unless marked, for the infinite lattice unless a size is given".
    • b0 and a1 are marked as the coefficients Ziff & Newman use, and the 3D Ising exponents as conformal-bootstrap estimates (Kos et al. 2016).
    • The header names every non-exact known value.
  • tests/critical.mjs: near()'s label comment calls V4 a numerically integrated value, not a measurement.
  • critical/model.js: V4 is marked as high-precision under "Exact results".
  • critical/README.md: the "Turbulent cascades" bullet links the turbulence study, in place of "This repository does not yet discuss turbulence elsewhere".
• 31d0f92
  • critical/README.md, "Turbulent cascades":
    • k^(−5/3) and the Re^(3/4) width are stated for air and water;
    • in the solar wind the ions' own motion, below which a steeper kinetic cascade carries on;
    • in molecular clouds the sonic scale ends only the supersonic range.
  • The short answer scopes the Reynolds number to an ordinary fluid. The root README says "the flow's own numbers ... set it".
  • critical/table.mjs: the open-edge shift is printed and described as leading order.
• 1ad7386
  • "Between the stirring and viscous scales" scopes k^(−5/3), and the width sentence stands on its own.
  • "below which" now follows the ion scales (about 100 km).
  • The short answer reads "(in an ordinary fluid)".
  • table.mjs says "Ziff & Newman's Table 1 values".

Not changed

• Critical-point study: wording fixes left over from #33's reviews #44's review also found two errors in Critical-point study: wording fixes left over from #33's reviews #44's description, and refuted a point about its rewrap. Neither needs a code change; Critical-point study: wording fixes left over from #33's reviews #44's description is corrected in place.
• Four refuted points stay as they are:
  • Re^(3/4) against the cascade model's Re^0.79;
  • a line break inside "1.6e-6 in U";
  • k^(−5/3) being stated for real air and water;
  • the root README's turbulence bullet, which already gives the solar-wind exception.

Checked (at 1ad7386)

• node tests/run.mjs passes all 8 check files on Node 18.20.8, 20.20.0 and 22.22.0.
• In tests/critical.mjs output, only the labels on the two Binder lines change.
• node critical/table.mjs prints every number identically to main; only labels and notes are added.
• python3 scripts/check_site.py and git diff --check are clean.

🤖 Generated with Claude Code

https://claude.ai/code/session_01XfgH6WcwTueUFZKb6fS2s2

[zero-aska]

Independent review of #45 at 7410d00, which is still the PR head. The review was done by agent instances that did not draft the branch: two reviewers (claims and sources; tests, prose and site), plus refuters for each finding (two for each blocking one).

What was verified

• The diff is one commit on 5fa2448 (main), +37/-30. It changes critical/README.md, critical/table.mjs, index.html and tests/critical.mjs. The PR body's account of each file matches the diff, except for the header claim in the blocking finding.
• Tests:
  • node tests/run.mjs passes all 8 check files on Node 18.20.8, 20.20.0 and 22.22.0.
  • CI (test 18/20/22) succeeded on 7410d00.
  • tests/critical.mjs gives 33 PASS and 0 FAIL at both base and head. The only output change is "exact" becoming "reference" on the two Binder lines.
  • The explicit d = 4 passed to near() is the same as fmt()'s default, so no digits change.
• node critical/table.mjs: only labels change in the output.
  • The heading changed, and so did the p_c and V4 notes.
  • The three p_c-based pooled rows now say "reference" (table.mjs:99-103). 91/48 is still "exact".
  • No number changed.
• Sources:
  • Salas and Sokal 2000 (arXiv:cond-mat/9904038). V4 = 1.1679229(47) is a CFT ratio of integrals over exact torus correlators, evaluated numerically: the numerator is a Monte Carlo integration with 10^9 points. The paper calls it "exact except for a numerical integration". So "high-precision, not exact" is right, and the stated error in U, 4.7e-6/3 = 1.6e-6, is correct.
  • Newman and Ziff 2000 (arXiv:cond-mat/0005264). p_c = 0.59274621(13) is a Monte Carlo estimate, so "measurement" or "reference" is the right label.
• critical/README.md:
  • In the table at :110-123, every row fits the new header "Exact or high-precision". The exact rows are Yang, Onsager, T_c, 1/8, 7/4, 1/4, 1/nu, xi and 91/48. The high-precision rows are the Binder value and p_c.
  • "lands on or near the known numbers" at :29 is accurate.
• No other line in the repository calls p_c or V4 exact. The one exception is the "Exact results" heading in critical/model.js (non-blocking item 3).
• index.html:238 matches the "Real cascades, and what stops them" table in turbulence/README.md: viscosity for the tunnel, surface layer and ocean, and the ions' own motion for the solar wind.
• Rewrapped comments stay within each file's existing width, and word diffs show only the intended edits. python3 scripts/check_site.py and git diff --check are clean.

Blocking findings

1. critical/table.mjs:10-11. The PR adds this header sentence: "The known values are exact except Newman & Ziff's p_c and Salas & Sokal's V4, which are high-precision values whose errors are negligible here". It is false, because the script prints other known values that are not exact. The PR body repeats the claim: "The header says which known values are not exact".
   • 3D Ising exponents. table.mjs:32-33 prints nu = 0.629971 and eta = 0.0362978 under "Exact and high-precision values (infinite lattice)" (:20). These are conformal-bootstrap estimates, 0.629971(4) and 0.0362978(20) (Kos, Poland, Simmons-Duffin and Vichi 2016; critical/README.md:297-298).
   • b0 and a1. table.mjs:29-31 prints b0 = 0.322 and a1 = 0.765 (from model.js:32-33). In Ziff and Newman 2002 (arXiv:cond-mat/0203496, sec. IV), a1 comes from earlier simulations and their own fit gives b0 = 0.320 +- 0.001. The open-edge shifts computed from them are printed beside the pooled p_c means (:102-103).
   • Ziff and Newman's Table 1. table.mjs:113-129 prints these measurements beside the nine-seed means. This is the weakest part, since line 6 already calls them measurements.
   • Fix. Name them all. For example: "exact except Newman & Ziff's p_c, Salas & Sokal's V4 and the 3D Ising exponents (high-precision), and Ziff & Newman's b0, a1 and Table 1 values (measurements)". Alternatively, write "exact except where marked" and label the 3D Ising and b0/a1 lines in the output. Either way, the PR body's bullet then becomes true.
   • Severity. The four refuters split. All four upheld the finding.
     • Both refuters on the tests track rated it blocking: it is new text, in a PR whose purpose is this labelling, and it is false however "known values" is read.
     • Both refuters on the claims track rated it non-blocking: it is a code comment, no number or printed comparison is wrong, and the wording on main was worse.
     • The blocking rating stands. The fix is one comment line.

Non-blocking findings

1. critical/table.mjs:20, 28-31. The heading "Exact and high-precision values (infinite lattice)" still covers b0 and a1, which are three-digit fitted coefficients. It also covers the open-edge shifts for L = 32/128 and 64/256, which are finite-size values. This predates the PR. Fix: label them, for example "(Ziff & Newman's coefficients)" after a1 and "(conformal bootstrap)" on the 3D Ising line. Alternatively, drop "(infinite lattice)".
2. tests/critical.mjs:53-55. The rewritten near() comment groups V4 under "a published measurement rather than an exact value". V4 is a CFT prediction evaluated by numerical integration. Salas and Sokal's measurements are the V4(L) values, which this file calls "measured V4(16)" at :174. The printed reference label is correct. Fix: "when it is a published measurement or high-precision value rather than an exact one".
3. critical/model.js:8-9, 21-23. The PR does not touch this file. Its header still lists V4 = 1.1679229(47) under "Exact results for the infinite lattice". It is the one place left that disagrees with the relabelling that the PR title says it finishes. Fix: "V4 = 1.1679229(47), a high-precision value (Salas & Sokal 2000)". Alternatively, head the list "Exact and high-precision results".
4. critical/README.md:269-272. The PR does not change these lines. They have been stale since turbulence/ was merged in #34.
   • "This repository does not yet discuss turbulence elsewhere" is false: turbulence/ is on main and linked from index.html:238.
   • "down to the scale at which viscosity removes it" holds only for ordinary fluids.
   • Fix: drop the sentence or link ../turbulence/. Optionally, qualify the viscosity clause with "in ordinary fluids such as air and water".
5. PR #45 body, "Not changed". It says "#44's review also noted three wording points in #44's description and its rewrap; it refuted one more". That miscounts. Beyond items 1-4, the review has two points about #44's description (the open-PR claim, and the review promised before merge) and one refuted point (the words the rewrap dropped). Fix: "#44's review also found two errors in #44's description and refuted a point about its rewrap."
6. PR #44 body (corrected), "Checked" bullet 3. It says "Four open PRs (#26, #35, #37, #39) change other parts of README.md". But #39 changes only notes/self-similarity.md, and #12, which is open, changes README.md at @@ -56,8 +56,10 @@. The review named #12, #26, #35 and #37. Fix: replace #39 with #12. The rest of the bullet then holds.
7. PR #44 body (corrected), paragraph before "Checked". It says "three words were also dropped ("it is", "those", "The")", which is four words. Fix: "four words" or "three phrases".
8. PR #45 body, footer. It ends with "Generated by [Claude Code](session URL)". The other PRs from this session (#33, #44) use a different form: a "Generated with Claude Code" line, then the session URL on its own line. Fix: edit the description.

Refuted:

• index.html:238 (the turbulence entry leaves out the steeper kinetic cascade below the solar wind's ion scale, and the molecular-cloud sonic scale). The entry describes the lower end of the constant-flux range, which matches turbulence/README.md:119. The sonic scale ends only the supersonic range, and the two media read as examples rather than a complete list.

Verdict: do not merge until the missing exceptions in critical/table.mjs:10-11 are fixed

---

Generated by Claude Code

[zero-aska]

60add3c addresses the review of 7410d00.

Blocking item (critical/table.mjs:10-11): fixed.

• The header now names every non-exact known value the script prints:
  • Newman & Ziff's p_c and Salas & Sokal's V4 (high-precision);
  • the 3D Ising exponents (conformal-bootstrap estimates, Kos et al. 2016);
  • b0 and a1 (the coefficients Ziff & Newman use);
  • their Table 1 values (measurements).
• Each is also marked where it is printed.
• The output heading now reads "Known values: exact unless marked, for the infinite lattice unless a size is given". This also covers non-blocking item 1: the open-edge shifts are finite-size values.

Non-blocking items

• Fixed in 60add3c:
  • critical/model.js marks V4 as high-precision under "Exact results".
  • near()'s label comment no longer calls V4 a measurement.
  • The "Turbulent cascades" bullet in critical/README.md says a scale set by the medium ends the cascade and links the turbulence study. Its "not yet discussed elsewhere" line is gone.
• Fixed in the PR descriptions:
  • Critical-point study: finish the exact/high-precision relabelling from #44's review #45's miscount of Critical-point study: wording fixes left over from #33's reviews #44's review;
  • Critical-point study: wording fixes left over from #33's reviews #44's open-PR list, which now reads Bound gravity-only witnesses for dark matter (issue 3) #12, Cosmic homogeneity: reproduce where galaxy clustering stops looking fractal, and fix the README's range #26, Halos within halos: how deep the nesting goes, and what the dark matter particle sets as the floor #35, Orbits: is a galaxy a big solar system? Where Kepler's law holds, where rotation flattens, and what ends a Kepler region #37;
  • Critical-point study: wording fixes left over from #33's reviews #44's word count, now four.
• Footer: the server rewrote this description's footer. It is now the standard form again.
• Left as is: the root README's "the flow's Reynolds number ... sets it". Nobody tunes the range in any medium, and critical/README.md now gives the detail.

Checks on 60add3c

• All 8 check files pass on Node 18, 20 and 22.
• critical/table.mjs prints every number identically to main. Only labels and notes are added.

An independent review of 60add3c is running.

(Edited: an earlier version called the open-edge-shift point "non-blocking item 2". It is item 1. The later re-review caught the mistake.)

---

Generated by Claude Code

[zero-aska]

Independent re-review of 60add3c (60add3c), the current head of #45. It was done by agent instances that did not draft the branch: two reviewers (one on the blocking item and claims, one on tests, prose and descriptions) and refuters for each finding.

The previous blocking item

Fixed. The critical/table.mjs:10-14 header now names every non-exact known value the script prints: p_c and V4 (high-precision), the 3D Ising exponents (bootstrap estimates), b0 and a1, and the Table 1 values (measurements). Each is marked where it is printed (:29-37). The pooled p_c rows print reference (:103-106), and Table 1 sits under "against Ziff & Newman's measurements". The one gap left is that the open-edge shift formula is truncated (non-blocking 2 below).

What was verified

• node tests/run.mjs passes all 8 check files on Node 18.20.8, 20.20.0 and 22.22.0. CI test (18), (20) and (22) succeeded on 60add3c. python3 scripts/check_site.py and git diff --check are clean. Comment widths stay within each file's existing maximum.
• node critical/table.mjs against main: every number is identical and only labels change (the only new numerals are the "4" in V4 and "2016"). tests/critical.mjs gives 33 PASS at both. Only the two Binder lines change, from exact to reference.
• Every value the script labels exact is exact: T_c, K_c, −√2, the 2D exponents, the ξ forms, 91/48, the exact power laws and the L^−5/48 fall. No "exact" label remains on a non-exact value in critical/README.md, critical/model.js, tests/critical.mjs, README.md or index.html.
• Sources:
  • Ziff & Newman 2002 (cond-mat/0203496) sec. IV uses b0 = 0.322 and a1 = 0.765, so "the coefficients Ziff & Newman use" is accurate.
  • Kos et al. 2016 (arXiv:1603.04436) give ν = 0.6299709 and η = 0.0362978.
  • Salas & Sokal 2000 (cond-mat/9904038) get V4 by integrating exact CFT correlators numerically. Its error in U is 1.6e-6.
• The PR body's per-file Changes match the diff, and "four non-blocking items" is right. #44's corrected description is right: four words, and the open PRs that touch README.md are #12, #26, #35 and #37.
• The fixes the reply describes are all present in 60add3c. The new ../turbulence/README.md link resolves. "Viscosity in air and water, the ions' own motion in the solar wind" matches turbulence/README.md:115-119 and index.html:238.

Blocking findings

1. critical/README.md:269-271: the k^(−5/3) sentence now covers media it does not describe.
   • The PR changed the cascade's lower end from "the scale at which viscosity removes it" to "a scale set by the medium", and the same bullet lists the solar wind and molecular clouds. So "In between, only the rate of energy flow matters, and the velocity spectrum falls as k^(−5/3)" now covers both.
   • The linked study contradicts that for both. turbulence/README.md:177-178 gives the solar-wind velocity spectrum as k^(−3/2) (Podesta, Roberts and Goldstein 2007). :220-227 give a dispersion slope of 0.49 ± 0.01 in the supersonic cloud range, where K41 gives 1/3.
   • The words predate the PR, but this scope is new. The PR scoped the next sentence to "an ordinary fluid" and left this one unscoped.
   • The claims reviewer raised this as non-blocking. Its refuter upheld it and rated it blocking, as a regression from a correctly scoped sentence. The other refuter, judging the same point inside non-blocking 1, called it non-blocking because the sentence follows "in air and water" and cites K41.
   • Fix: "In air and water, only the rate of energy flow matters, and the velocity spectrum falls as k^(−5/3)". Or use "in an ordinary fluid" to match the next sentence.

Non-blocking findings

1. critical/README.md:273-275: the bullet drops the qualifiers on the sonic scale and the ion scale. "Lists what ends real cascades: ... the ions' own motion in the solar wind, the sonic scale in molecular clouds" reads as if the cloud cascade stops at the sonic scale. The study says only the supersonic range ends there and a shallower subsonic cascade continues (turbulence/README.md:121, :193-195). README.md:34-36 words it the same way. The solar-wind clause drops "(the MHD range)" in the same way, though a steeper kinetic cascade continues there (turbulence/README.md:119, :156-158). Both reviewers found this. Fix: follow README.md:34-36, for example "viscosity in air and water; in the solar wind the ions' own motion ends the MHD range, and in molecular clouds the sonic scale ends the supersonic range, with other cascades continuing below both".
2. critical/table.mjs:32 and :106: the open-edge shifts are printed under "exact unless marked" with no mark of their own.
   • The formula is only the leading term of Ziff & Newman's expansion; model.js:321 says "to leading order". Their measured crossings sit at 0.862, 0.930, 0.965 and 0.975 of the shift at L = 32, 64, 128 and 256.
   • The header list at :10-14 does not mention the truncation, and the pooled lines print "(open-edge shift +0.00040)" with no mark.
   • No number is wrong. The truncation moves the shifts by less than a quarter of the error between seeds.
   • The reply answers the previous review's item 1 with "unless a size is given". That covers finite size, not the truncation.
   • Fix: add "to leading order" at :32 (and optionally at :106). Name the shifts in the header, for example "the open-edge shifts are the leading term of Ziff & Newman's expansion, from those coefficients".
3. critical/README.md:43-45 and README.md:355-357: the summaries say the Reynolds number sets the range. The bullet they summarise now limits Re^(3/4) to "an ordinary fluid" and names cut-offs that are not viscous. turbulence/README.md:154 also says the solar wind is too dilute for viscosity to end its cascade. The PR body's reason for keeping README.md:356 ("Nobody tunes the range in any medium") supports "nobody tunes", not "the Reynolds number sets it". Both lines predate the PR. Fix, in both places: "through the Reynolds number in ordinary fluids", or "the flow's own numbers".
4. PR body, Status bullet 1: "7410d00 made the changes below" is wrong.
   • Six items under Changes exist only in 60add3c:
     • the model.js V4 marker;
     • the Turbulent cascades rewrite and its link;
     • the "Known values" heading;
     • the b0/a1 and 3D Ising markers;
     • the header naming every non-exact value;
     • the near() comment.
   • The same bullet says 7410d00's header "named only p_c and V4", yet the Changes list attributed to 7410d00 says the header names all the non-exact values.
   • Fix: "The two commits made the changes below", or say which commit made each item.
5. Reply comment 6462067284, Blocking item, last bullet: "non-blocking item 2" should be "item 1". The point about the heading and the open-edge shifts was item 1 of review 5678462188. Item 2 was the near() comment, which the reply also addresses separately. Both reviewers found this. Fix: "non-blocking item 1".

Refuted:

• critical/README.md:271-274 (Re^(3/4) against the shell model's Re^0.79): the sentence is about ordinary fluids. The study sizes real flows with 3/4 itself and says its shell model is more intermittent than measured flows. Alexakis and Biferale 2018, eq. (41), give 3/4.
• tests/critical.mjs:54-55 (the rewrap separates "1.6e-6" from "in U"): the rewrap follows the file's habit of filling lines to about 120 columns. Lines 28-29 on main already separate a number from its qualifier in the same way.

Verdict: do not merge until the blocking items (the k^(−5/3) scope at critical/README.md:269-271) are fixed

---

Generated by Claude Code

[zero-aska]

31d0f92 addresses the re-review of 60add3c.

Blocking item (critical/README.md:269-271, the k^(−5/3) scope): fixed. The bullet now reads: "In air and water that is the scale at which viscosity removes it; in between, only the rate of energy flow matters, the velocity spectrum falls as k^(−5/3) ..., and the flow sets the range itself: its width grows as the Reynolds number to the power 3/4". So both the spectrum and the Re^(3/4) width are stated for air and water only.

Non-blocking items

• 1, the sonic scale and the ion scale. The bullet now says the sonic scale "ends only the supersonic range" in molecular clouds, and that a steeper kinetic cascade "carries on" in the solar wind. That is turbulence/README.md's word; the root README says "takes over".
• 2, the open-edge shift. critical/table.mjs now prints and describes it as leading order: "to leading order, open edges put the crossing ...", "leading-order open-edge shift". The header says the shifts keep only the leading term.
• 3, the Reynolds-number summaries.
  • critical/README.md's short answer now says "through the Reynolds number in an ordinary fluid".
  • The root README now says "the flow's own numbers ... set it".
  • My earlier reply's reason for leaving the root README alone did not support its wording, as the re-review noted.
• 4, the PR body's attribution. Fixed: the body now gives each commit its own items.
• 5, the wrong item number in my earlier reply. Fixed in place, with a note.
• Refuted, so unchanged: Re^(3/4) against the shell model's Re^0.79, and the line break inside "1.6e-6 in U".

Checks on 31d0f92

• All 8 check files pass on Node 18, 20 and 22.
• critical/table.mjs prints every number identically to main.
• scripts/check_site.py and git diff --check are clean.

(Edited after the review of 31d0f92, which caught three slips:

• the item numbers now follow the published re-review, and internal reviewer labels are removed;
• item 1 no longer says the line names the ion scales: at 31d0f92 it did not, and 1ad7386 adds them;
• item 1 no longer says "carries on" is the root README's word.

The wording fixes themselves are in 1ad7386.)

---

Generated by Claude Code

[zero-aska]

This is an independent re-review of 31d0f92 (31d0f92), which is still the PR head. It was done by agent instances that did not draft the branch. Two reviewers covered (a) the blocking item and claims and (b) tests, prose and descriptions. Every finding then went to a refuter. No finding was raised as blocking, so none needed a second refuter.

The previous blocking item

It is fixed. In critical/README.md:269-275, the k^(−5/3) spectrum and the Re^(3/4) width now sit in a sentence that opens "In air and water that is the scale at which viscosity removes it; in between, ...". Neither is said of the solar wind or of molecular clouds any more. The next sentence gives each medium its own cut-off: viscosity in air and water, the ions' own motion in the solar wind, and the sonic scale in molecular clouds, which ends only the supersonic range. This matches turbulence/README.md:115-121, :154-158 and :193-195. It no longer contradicts the solar-wind k^(−3/2) spectrum (:177-178) or the 0.49 supersonic slope (:220-223).

What was verified

• node tests/run.mjs passes all 8 check files on Node 18.20.8, 20.20.0 and 22.22.0. CI test (18), (20) and (22) succeeded on 31d0f92.
• tests/critical.mjs gives 33 PASS and 0 FAIL both on main and at 31d0f92. Only the two Binder lines change, from exact to reference.
• node critical/table.mjs prints every number identically to main and to 60add3c. Between 60add3c and 31d0f92, only lines 8, 61 and 62 differ, and all three are labels.
• The "leading order" wording matches Ziff & Newman 2002 (arXiv:cond-mat/0203496), eq. (27):
  • the repository's shift reduces to their leading term at L1 = L/2;
  • the next terms are of relative order L^(−ω), with ω = 0.90 ± 0.02;
  • b0 = 0.322 and a1 = 0.765 are their sec. IV values.
• No touched file puts an exact label on a non-exact value. p_c, V4, the 3D Ising exponents, b0/a1, the open-edge shifts and the Table 1 values are each marked.
• The re-review's non-blocking items 1-3 are applied (sonic scale, open-edge shift, Reynolds-number summaries). critical/README.md:43-45 and README.md:356 are accurate. No touched file still has a Reynolds-number or k^(−5/3) statement without a scope.
• Edited lines stay within each file's existing width. Word-sequence diffs of every rewrapped paragraph show only the intended edits, and no value or link text is split.
• python3 scripts/check_site.py and git diff --check origin/main...31d0f92 are clean. The ../turbulence/README.md link resolves, and the #tuned-scale-invariance-against-self-organised-ranges anchor exists (critical/README.md:262).
• The PR body's per-commit Changes match git show of each commit, and "No number, tolerance or check changes" holds. Reply 6462067284 now says "non-blocking item 1" and carries an edit note.

Blocking findings

None.

Non-blocking findings

1. critical/README.md:270-272: "in between," now governs three clauses. The third clause, "the flow sets the range itself: its width grows as the Reynolds number to the power 3/4", describes the whole range, not the scales inside it. 31d0f92 also dropped "in an ordinary fluid" from the width clause. The Re^(3/4) scope now depends on "in between" carrying "In air and water" across the semicolon. The short answer at :44 is now more explicit than the bullet it summarises.
   • Fix: "In air and water that is the scale at which viscosity removes it. Between the stirring and viscous scales only the rate of energy flow matters, and the velocity spectrum falls as k^(−5/3) (Kolmogorov 1941). The flow sets that range itself: its width grows as the Reynolds number to the power 3/4 ...".
   • Alternatively, put "in an ordinary fluid" back on the width clause, as at 60add3c.
2. critical/README.md:274: in "in the solar wind the ions' own motion, below which a steeper kinetic cascade carries on", "below which" can only refer to "the ions' own motion". That is a mechanism, not a scale. turbulence/README.md:119 keeps the scale ("the ion scales, about 100 km") separate from the cause.
   • The PR body and reply 6462528933 both describe the line as saying "below the ion scales", but the line never names them.
   • The reply also says "carries on" is how the root README puts it. README.md:34 says "takes over"; only turbulence/README.md:157 uses "carries on".
   • Fix: "in the solar wind the ions' own motion, at the ion scales (about 100 km), below which a steeper kinetic cascade carries on". Rewrap within 115 columns.
3. critical/README.md:44-45: "through the Reynolds number in an ordinary fluid or the time gravity has had to act" is a garden path. "In ... the time gravity has had to act" reads as one phrase until the reader backtracks. The intended split is still the only sensible one.
   • Fix: "through the Reynolds number (in an ordinary fluid) or the time gravity has had to act", or put a comma before "or". README.md:356 already uses commas this way.
4. critical/table.mjs:13-14: "their Table 1 values" now follows "the open-edge shifts built from them" instead of Ziff & Newman. Read as b0's and a1's values, the sentence is wrong, because those coefficients come from sec. IV (critical/model.js:32-33).
   • Fix: "and Ziff & Newman's Table 1 values are measurements". This brings line 14 to 83 of 120 columns.
5. Reply 6462528933 numbers the re-review's items wrongly. This is the same kind of error as the re-review's non-blocking 5, which the reply says it fixed.
   • It calls the PR-body attribution "2"; in review 5678702151 it is 4. It calls the earlier reply's item number "3"; it is 5.
   • It tags items "claims N3", "claims N4" and "claims N5". Those labels appear nowhere in the published review. There, the open-edge shift and Reynolds summaries are 2 and 3, and the Re^0.79 point is unnumbered.
   • Fix: edit the reply to use 2, 3, 4 and 5, drop the internal labels, and refer to the refuted point by its text.

Refuted:

• k^(−5/3) stated as fact for real air and water (critical/README.md:270-271). The wording was already on main, and this commit only makes its scope explicit. It is the fix the last re-review asked for. The study itself reports measured k^(−5/3) spectra (turbulence/README.md:144-145, :232-233), and its intermittency figure moves the slope by only about 1.6%.
• Root README.md:29-30 still opens "Between the size where a flow is stirred and the size where viscosity takes over". This predates the PR. The same bullet gives the solar-wind exception at :33-34, which is what the #33 and #44 reviews said the site entry lacked.

Verdict: merge

---

Generated by Claude Code

[zero-aska]

The review of 31d0f92 said merge. It found four wording slips that 31d0f92 itself introduced, so I fixed them before merging rather than leave them on main. 1ad7386 applies them as worded:

1. "in between" governed three clauses.

   • critical/README.md now reads "Between the stirring and viscous scales only the rate of energy flow matters, and the velocity spectrum falls as k^(−5/3) (Kolmogorov 1941)".
   • The width clause is its own sentence: "The flow sets that range itself: its width grows as the Reynolds number to the power 3/4".

2. "below which" hung on a mechanism. It now reads "in the solar wind the ions' own motion, at the ion scales (about 100 km), below which a steeper kinetic cascade carries on".

3. The short answer was a garden path. It now reads "through the Reynolds number (in an ordinary fluid) or the time gravity has had to act".

4. critical/table.mjs header. It now says "Ziff & Newman's Table 1 values", not "their".

5. My reply 6462528933 had three slips. It is corrected in place, with a note saying what changed:

   • It numbered the items wrongly and used internal reviewer labels.
   • It said the line named the ion scales, which it did not at 31d0f92.
   • It credited "carries on" to the root README.

The two refuted points stay as they are.

All 8 check files pass on Node 18, 20 and 22. scripts/check_site.py and git diff --check are clean. 1ad7386 changes prose and one comment only. A short independent check of 1ad7386 follows before merge.

---

Generated by Claude Code

[zero-aska]

This is a short independent check of the delta 31d0f92..1ad7386. It was done by agent instances that did not draft the branch: one reviewer, plus one refuter per finding (two for blocking ones). Review 5678939490 of 31d0f92 covered everything before this delta.

What was verified

• Scope. git diff --stat 31d0f92 1ad7386 touches only critical/README.md (+9/-8) and critical/table.mjs (+1/-1). git diff --check is clean for both 31d0f92..1ad7386 and 5fa2448..1ad7386. The PR as a whole is still 6 files, +56/-42, which matches its metadata.
• The four fixes match what review 5678939490 asked for.
  • critical/README.md:270-272 closes the "in between" clause chain: "Between the stirring and viscous scales only the rate of energy flow matters ... its width grows as the Reynolds number to the power 3/4". The Re^(3/4) sentence now covers only the stirring-to-viscous range.
  • critical/README.md:274-275: "the ions' own motion, at the ion scales (about 100 km), below which a steeper kinetic cascade carries on". "below which" now refers to a scale. This agrees with turbulence/README.md:119 and :156-157 and with the root README.md:33-34.
  • critical/README.md:44-45: "(in an ordinary fluid)" is now in parentheses.
  • critical/table.mjs:14: "and Ziff & Newman's Table 1 values are measurements".
• Origin of the slips. git diff 60add3c 31d0f92 confirms that all four were introduced in 31d0f92, as the new reply says.
• Nothing lost in the rewrap. The word sequences of the three rewrapped passages differ only by the review's edits. Edited Markdown lines are at most 115 characters, the file's existing maximum. table.mjs:14 is 83 columns. No value, citation or link text is split across lines. Anchors used by other files are unchanged.
• Checks. At 1ad7386, node tests/run.mjs passes on Node 22.22.0 and 20.20.0 ("all 8 check file(s) pass"), and python3 scripts/check_site.py prints "site check clean". CI run 40647938334 passed test (18), test (20) and test (22).
• Replies, PR body and commit message.
  • Edited reply 6462528933 now follows review 5678702151's numbering (items 1-5). It names the refuted points by their text and correctly attributes "carries on" to turbulence/README.md and "takes over" to the root README.
  • New reply 6462920014 quotes all four fixes accurately.
  • The PR body's 1ad7386 bullets and Status bullet, and the commit message, all match the diff.

Blocking findings

None.

Non-blocking findings

1. The PR body still credits the ion-scale wording to 31d0f92 (PR body, "Changes, by commit" > 31d0f92). That bullet still reads "in the solar wind a steeper kinetic cascade carries on below the ion scales". At 31d0f92, critical/README.md:274 read "in the solar wind the ions' own motion, below which a steeper kinetic cascade carries on". It did not name the ion scales; only 1ad7386 adds "at the ion scales (about 100 km),", and the body's own 1ad7386 bullet says so. This is the PR-body half of review 5678939490's non-blocking 2, and it is still open. Fix: change the 31d0f92 bullet to "in the solar wind the ions' own motion, below which a steeper kinetic cascade carries on".
2. The edit note on reply 6462528933 does not mention that the reply's content changed. Review 5678939490 (non-blocking 2) quotes the original reply as saying the line names the ion scales and that "carries on" is how the root README puts it. The current item 1 says neither and now gives the opposite attribution, which is correct ("the root README says 'takes over'"). The edit note mentions only that the item numbers now follow the published re-review and that internal reviewer labels are removed. Item 5 of reply 6462920014 says the same. A reader comparing the review's quotes with the current reply has nothing that explains the change. Fix: extend the edit note, and item 5 of 6462920014, with something like "and item 1 no longer says the line names the ion scales or that 'carries on' is the root README's word".

No findings were refuted.

Verdict: merge

---

Generated by Claude Code