---
slug: flat-out-of-nothing
passed: 2026-10-11
by: claude-vibrant-lamport-79znpp, two independent Claude checkers on oversight/claims-pass.md (source first, every claim the page makes), an adjudicator who re-read each finding at its source and a skeptic who tried to make the page right; every HIGH and every lone WRONG re-read by the instance at the source before fixing; fixes made by a fixer agent and read against the sources by a separate fix-checker
checked: 51
confirmed: 36
wrong: 10
unverifiable: 2
observed: 3
stale: 0
---
# Claims pass: Flat Out of Nothing

Reader-side check, run before the fixes from an empty directory by both checkers and the adjudicator with the same result: `curl -sS -L -o verify-flat-out-of-nothing.mjs https://artwaste.land/checks/research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs` returned HTTP 200, 14,887 bytes, byte-identical to research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs, and `node verify-flat-out-of-nothing.mjs` (node v22.22.0) gave exit 0, 27 ok, 0 FAIL, "ALL CHECKS PASSED" (A's run of the literal repository-path recipe gave the same). The figures it printed matched the page: R = 1.56 km, 1/R = 6.40e-4 /m, gap RMS 0.00 µm at 100% contact, determinant 2, convergence in 9 cycles, 0 nm recovery error. The self-reading probe (step 6) does not apply: the verifier opens no file (grep for import, require(, readFile, process.argv and fetch( finds nothing) and the page prints no transcript of its output. B showed that an altered copy of the page ("R = 9.99 km", determinant 3) still passes, as a script that never reads the page must; the aside says so, and it bears only on the wording "re-checked offline" (F8). After the fixes, which changed only the verifier's comments, the repository verifier still gives 27 ok, 0 FAIL, exit 0.

Three claim lines are WRONG and 7 MINOR; the MINOR lines are counted as wrong. All nine merged findings were fixed. The two WRONG findings answer the three WRONG lines (F1 answers two and has a findings line for each): the page said the laser never escaped the bootstrap, which holds only for the usual calibration, since a liquid surface levelled by gravity (in 2002 Germany's national primary flatness standard was a mercury mirror) and PTB's angle-measuring method are absolute references outside the flats (F1); and Schulz and Schwider's own 1967 method already reached any diameter from four interferograms, so going beyond one diameter did not wait for Fritz in 1984 (F2). The seven MINOR findings: a sentence credited to the wrong Wikipedia article (F3), Wikisource said to give neither venue nor date (F4), the British Association's own title for the paper left out (F5), three slips in the layer's README (F6), Maudslay's precedence stated flatly with no source (F7), "re-checked offline" for a verifier that does not read the page (F8), and a run instruction with no link (F9). The page now carries a dated "Corrected 2026-10-11" line for F1 and F2. The two UNVERIFIABLE lines (what the paywalled Zhai et al. survey says, and shop guides contrasting rotated pairings with one fixed sequence) carry no finding; the page no longer says what the survey states. No finding was declined and none is held by a note.

## Claims
- CONFIRMED | Joseph Whitworth systematized the answer in 1840 ... publish it, in an 1840 paper read to the British Association at Glasgow | https://archive.org/download/miscellaneouspa00whitgoog/miscellaneouspa00whitgoog_djvu.txt (Whitworth, Miscellaneous Papers on Mechanical Subjects, 1858: Preface and half-title) | The paper upon plane metallic surfaces was read at the British Association in Glasgow, in the year 1840 ... / A PAPER ON PLANE METALLIC SURFACES OR TRUE PLANES, READ AT THE MEETING OF THE BRITISH ASSOCIATION IN GLASGOW, 1840.; The paper was read in 1840 and first printed in 1841 (Mechanics' Magazine; BA Report). 'Published 1840' is fair if read as 'made public'.
- CONFIRMED | reprinted in his collected papers as A Paper on Plane Metallic Surfaces or True Planes | https://archive.org/download/miscellaneouspa00whitgoog/miscellaneouspa00whitgoog_djvu.txt (1858 Contents) | A Paper on Plane Metallic Surfaces or True Planes . . 1
- CONFIRMED | Grace's Guide reports it as "Plane Metallic Surfaces, and the Proper Mode of Preparing them" | https://www.gracesguide.co.uk/Joseph_Whitworth | In 1840, at the meeting of the British Association, held in Glasgow, Whitworth read a Paper on 'Plane Metallic Surfaces, and the Proper Mode of Preparing them,'
- CONFIRMED | the 1841 Mechanics' Magazine printing carries that wording too | https://archive.org/download/dli.bengal.10689.16198/10689.16198_djvu.txt (Mechanics' Magazine vol. 34, 1841, p. 39); https://circuitousroot.com/artifice/machine-shop/surface-finishing/hand-scraping/index.html | ON PLANE METALLIC SURFACES, AND THE PROPER MODE OF PREPARING THEM; READ AT THE MEETING OF THE BRITISH ASSOCIATION, AT GLASGOW, A.D, 1840, BY JOSEPH WHITWORTH, ESQ. (CircuitousRoot: Volume 34, No. 910 (Saturday, January 16, 1841), pp. 39-42)
- MINOR | (The exact title varies by printing: Grace's Guide reports it as ... and the 1841 Mechanics' Magazine printing carries that wording too.) / for ... the title as read there, Grace's Guide | https://archive.org/download/reportofbritisha40brit/reportofbritisha40brit_djvu.txt (BA Report 1840, Notices and Abstracts p. 192) | On producing True Planes or Surfaces on Metals. By Mr. Joseph Whitworth.; F5. The two titles given are accurate; the BA's own third title is omitted.
- MINOR | Joseph Whitworth did not discover it alone: the practice already lived in Henry Maudslay's London workshop (also: 'The three-plate method predates Whitworth ... notably Henry Maudslay's'; relates note 'he systematized a procedure the workshop already held') | https://archive.org/download/jamesnasmythengi00nasmiala/jamesnasmythengi00nasmiala_djvu.txt (Nasmyth 1883, ch. VIII, pp. 148-149); https://en.wikipedia.org/w/index.php?title=Surface_plate&action=raw (citing Rolt 1965) | Three of each were made at a time, so that by the mutual rubbing of each on each the projecting surfaces were effaced ... This art of producing absolutely plane surfaces is, I believe, a very old mechanical "dodge." / Surface plate: He originated the systems of scraping a cast-iron plate to flatness ... and of working plates in sets of three; F7. Supported, but no source is given on the page and the history is contested. Grace's Guide (obituary): 'while here he made his first great discovery'. Nasmyth arrived at Maudslay's in 1829.
- CONFIRMED | he argued for finishing by hand scraping rather than grinding, so that the high spots shown by the colouring matter could be taken down one at a time | https://en.wikisource.org/wiki/Miscellaneous_Papers_on_Mechanical_Subjects/A_Paper_on_Plane_Metallic_Surfaces_or_True_Planes | To obtain any such result, it is necessary to possess the means of operating independently on each point, as occasion may require, whereas grinding affects all simultaneously ... afterwards shows the prominences to be removed by the scraping instrument
- CONFIRMED | Whitworth used red ochre and oil for that transfer | https://en.wikisource.org/wiki/Miscellaneous_Papers_on_Mechanical_Subjects/A_Paper_on_Plane_Metallic_Surfaces_or_True_Planes ; Mechanics' Magazine 1841, https://archive.org/download/dli.bengal.10689.16198/10689.16198_djvu.txt | Colouring matter, such as red ochre and oil, is spread over the surface plate, as equally as possible; 'such as': red ochre is Whitworth's example. The BA's 1840 abstract has 'red chalk and oil' (per checker A).
- CONFIRMED | Prussian-blue engineer's blue is the later equivalent | https://www.gracesguide.co.uk/Joseph_Whitworth; https://en.wikipedia.org/w/index.php?title=Prussian_blue&action=raw | using colouring matter (the forerunner of today's 'engineer's blue') / Engineer's blue, Prussian blue in an oily base, is the traditional material used for spotting metal surfaces such as surface plates
- CONFIRMED | "While grinding is universally regarded as indispensable to a finished surface, it is, in fact, positively detrimental." | https://en.wikisource.org/wiki/Miscellaneous_Papers_on_Mechanical_Subjects/A_Paper_on_Plane_Metallic_Surfaces_or_True_Planes ; Mechanics' Magazine 1841 p. 39, https://archive.org/download/dli.bengal.10689.16198/10689.16198_djvu.txt | While grinding is universally regarded as indispensable to a finished surface, it is, in fact, positively detrimental.; Word for word in both the 1841 and 1858 printings.
- CONFIRMED | work three plates against each other in pairs | https://en.wikisource.org/wiki/Miscellaneous_Papers_on_Mechanical_Subjects/A_Paper_on_Plane_Metallic_Surfaces_or_True_Planes | Take three plates of cast iron ... 1st. Nos. 2 and 3 to No. 1. 2nd. Nos. 2 and 3 to each other. 3rd. No. 1 to Nos. 2 and 3. These parts compose an entire series, by repeating which a gradual approach is made to absolute truth
- CONFIRMED | One can be a shallow dome and the other its exact bowl ... no test between the two can pin down; the third plate kills the free parameter | https://www.gracesguide.co.uk/Joseph_Whitworth (Whitworth's own account, quoted in the obituaries) | But the coincidence of two surfaces would not prove them to be planes. If one were concave and the other convex, they might still coincide. I got over this difficulty by taking a third surface, and adjusting it to both of the others.
- CONFIRMED | The three-plate method needs no prior flat reference. It is the answer to the bootstrapping question of the first flat surface. | https://en.wikipedia.org/w/index.php?title=Flatness_(manufacturing)&action=raw | This method does not rely on other flat reference surfaces or other precision instruments, and thus solves the bootstrapping problem of how to create the first precise flat surface.
- CONFIRMED | A Fizeau interferometer does not measure a flat against the truth. It measures a flat against its own reference flat, and reads out the gap. | https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=823338 (Griesmann et al., Opt. Eng. 46(9) 2007) | the combined flatness error of test and reference wavefronts is measured ... W(x,y) = WR(-x,y) + WT(x,y)
- WRONG | The laser did not escape the bootstrap ... To learn the reference you are back to comparing flats against each other, which is why calibration is again three flats measured pairwise (dek/meta: 'The modern laser did not escape this ... so calibration is again three flats compared pairwise'; og: 'The laser never escaped the bootstrap.') | https://opg.optica.org/ao/abstract.cfm?uri=ao-6-6-1077 ; https://www.ptb.de/cms/en/presseaktuelles/journals-magazines/ptb-news/ptb-news-ausgaben/archivederptb-news/news02-3/ultraprecise-flatness-measurements-of-optical-surfaces.html ; https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=823338 | This absolute determination can be effected in connection with a liquid surface, or (as done by the authors) only by suitable evaluation of relative measurements between unknown plates in various positional combinations. / Until now, the actual deviation of a real surface from perfect flatness is determined by comparison with a liquid mercury mirror which is the national primary standard for flatness in Germany. / full-aperture three-flat tests have largely superseded earlier calibration methods such as liquid-mirror flatness standards.; F1. The three-flat test is the usual route, not the only one.
- WRONG | every reference flat is still calibrated against other flats, and the bootstrap is closed, never escaped | https://api.crossref.org/works/10.4028/www.scientific.net/kem.381-382.543 (Geckeler, PTB, 2008); https://www.ptb.de/cms/en/presseaktuelles/journals-magazines/ptb-news/ptb-news-ausgaben/archivederptb-news/news02-3/ultraprecise-flatness-measurements-of-optical-surfaces.html; https://api.openalex.org/works/doi:10.1364/AO.37.002579 | The ESAD method is optimally suited for creating a primary standard for straightness and flatness with highest accuracy by which the three-flat test or liquid mirrors can be replaced as starting points of the traceability chain in flatness measurement. / it does not require an external flatness standard. Instead, the method utilizes the straight propagation of the light beam as a reference. / Powell & Goulet 1998: a variation of the liquid-flat technique for determining the absolute flatness of a 240-mm-diameter optical surface; F1.
- CONFIRMED | calibration is again three flats measured pairwise. Schulz and Schwider set this out in 1967, and its three-body structure is Whitworth's exactly | https://opg.optica.org/ao/abstract.cfm?uri=ao-6-6-1077 ; https://wp.optics.arizona.edu/jcwyant/wp-content/uploads/sites/13/2016/08/10-Absolute-Measurements.pptx.pdf | (as done by the authors) only by suitable evaluation of relative measurements between unknown plates in various positional combinations / For an absolute measurement of flat surfaces need at least three flats – Compare flat A with flat B – Compare flat A with flat C – Compare flat B with flat C; Schulz and Schwider do describe the pairwise route, but the same sentence names a liquid surface as the alternative (see F1).
- CONFIRMED | G. Schulz and J. Schwider, "Precise measurement of planeness", Applied Optics 6(6), 1077–1084 (1967) | https://api.crossref.org/works/10.1364/AO.6.001077 ; https://opg.optica.org/ao/abstract.cfm?uri=ao-6-6-1077 | Precise Measurement of Planeness; G. Schulz, J. Schwider; Applied Optics; 6; 6; 1077; 1967-06-01 (Optica meta: lastpage 1084)
- WRONG | the result that the classical test yields the absolute profile along one diameter only (owing to the flip inversion): G. Schulz and J. Schwider ... (1967). The rotation and Zernike extension: B. S. Fritz / the null space that the rotation extensions (Fritz, 1984) exist to pin | https://opg.optica.org/ao/abstract.cfm?uri=ao-6-6-1077 ; https://wp.optics.arizona.edu/jcwyant/wp-content/uploads/sites/13/2016/08/Absolute-testing-of-flats.pdf | an absolute method which determines the deviations of a surface from ideal planeness along arbitrary central sections, without a liquid surface, from four relative interference photographs. / In the traditional three-flat method, the flats are compared in pairs. By rotating the flats with respect to each other, the exact profiles along several diameters of each flat are obtained.; F2 (PARTLY, MED). The one-line limit of three comparisons is real (Griesmann 2006) and is commonly cited to Schulz & Schwider 1967 (Uchikoshi et al. 2013). What is wrong is the chronology: Schulz & Schwider 1967 already reached any diameter with a fourth measurement, and Fritz's addition was the full-aperture Zernike fit.
- CONFIRMED | Along the one line about which a flat is flipped ... the three comparisons are clean sums ... That 3-by-3 system solves ... The classical three-flat test recovers the absolute profile along one diameter and no more | https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=822358 (Griesmann, Appl. Opt. 45, 5856, 2006) | Equation (2) can be solved only for the vertical line at x = 0 because, in the interferometer coordinate system, WB(x, y) and WB(-x, y) must be considered different variables ... A solution can only be found if at least one additional measurement is made.; True of three pairwise comparisons; see F2 for the attribution.
- CONFIRMED | Off the flip axis the flip mirrors the surface, x → -x, so the comparison reads A(x) + B(-x). Add the same odd bow to all three flats and every measurement is untouched | https://wp.optics.arizona.edu/jcwyant/wp-content/uploads/sites/13/2016/08/AbsoluteTestingOfFlatsDecomposedToEvenAndOddFunctions.pdf ; Griesmann et al. 2007 Eq. (1), https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=823338 ; research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; recreated: node verify-flat-out-of-nothing.mjs | If two flats are F(x,y) and G(x,y), and G(x,y) is flipped in x, then the measured OPD is equal to F(x,y)+G(-x,y). / verifier: max/delta measurement/ = 0 nm
- CONFIRMED | Fritz turned the flats and fit Zernike polynomials in 1984 | https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=822358 ; https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=823338 | Parks and later Fritz have described a method of solving Eq. (39) in which ΩA is described by a (truncated) Zernike series. / In the flat test solutions of Parks and Fritz14 ΩA and Δ are expressed as Zernike series; True on its own. The page does not claim Fritz was first; for the chronology see F2.
- CONFIRMED | B. S. Fritz, "Absolute calibration of an optical flat", Optical Engineering 23(4), 379–383 (1984) | https://api.crossref.org/works/10.1117/12.7973304 | Absolute Calibration Of An Optical Flat / Optical Engineering / 23 / 4 / 379-383 / 1984-08-01
- CONFIRMED | the N-position and shifted-map methods since do the same. Every one of them is another mutual comparison. None escapes the bootstrap | https://opg.optica.org/ao/abstract.cfm?uri=ao-35-7-1015 (Evans & Kestner 1996); https://opg.optica.org/ao/abstract.cfm?uri=ao-37-25-5951 (Parks, Shao, Evans 1998) | if either the test part or the reference surface in an interferometric test is rotated to N equally spaced positions ... It is also shown how this method may be applied in an absolute test; 'None escapes' refers to these mutual-comparison methods only, so it is true. The universal version is F1.
- CONFIRMED | Pixel-based absolute surface metrology by three flat test with shifted and rotated maps (ScienceDirect) ... Both are surveyed | https://api.crossref.org/works/10.1016/j.optlaseng.2017.10.021 | title 'Pixel-based absolute surface metrology by three flat test with shifted and rotated maps'; Zhai, Chen, Xue, Yin; Optics and Lasers in Engineering 102, 92-99 (2018); alternative-id S0143816617306188; bib0004 Schulz 'Precise measurement of planeness' 1967; bib0006 Fritz 'Absolute calibration of an optical flat' 1984; It cites both. Its references 2-3 are liquid-flat references (Powell & Goulet 1998; Vannoni & Molesini 2005).
- UNVERIFIABLE | with the one-diameter limitation stated, in Pixel-based absolute surface metrology ... | not kept: abstract and text of 10.1016/j.optlaseng.2017.10.021 (ScienceDirect page https://www.sciencedirect.com/science/article/abs/pii/S0143816617306188, linked from the page footer, returned 403 to curl and WebFetch; https://api.crossref.org/works/10.1016/j.optlaseng.2017.10.021 has abstract None; Semantic Scholar has the abstract elided by the publisher) | Crossref 'abstract': None; Semantic Scholar: "The following paper fields have been elided by the publisher: {'abstract'}"; Positive control: the same tools retrieved the Optica abstract of Schulz & Schwider.
- MINOR | Wikisource ... which carries the text but states neither venue nor date | https://en.wikisource.org/wiki/Miscellaneous_Papers_on_Mechanical_Subjects/A_Paper_on_Plane_Metallic_Surfaces_or_True_Planes ; https://archive.org/download/miscellaneouspa00whitgoog/miscellaneouspa00whitgoog_djvu.txt (Preface) | That is, prior to the year 1840. / This, it must be borne in mind, had reference to the state of things existing in 1840. / Preface: The paper upon plane metallic surfaces was read at the British Association in Glasgow, in the year 1840; F4. The venue is absent from the chapter page; the date appears in its footnotes; the same edition's Preface gives both.
- MINOR | Surface plate (Wikipedia), which states the method needs no other flat reference and solves the bootstrapping problem of the first flat surface | https://en.wikipedia.org/w/index.php?title=Surface_plate&action=raw ; https://en.wikipedia.org/w/index.php?oldid=1287018505&action=raw | Unlike most mechanical precision instruments, surface plates do not derive their precision from more-precise standards. Instead they originate precision by application of the principle of "automatic generation of gages".; F3. 'bootstrap' appears 0 times in either revision; the sentence is in Flatness (manufacturing).
- CONFIRMED | the two-plate concave/convex failure mode and Whitworth's 1830s popularization: Flatness (manufacturing) (Wikipedia) | https://en.wikipedia.org/w/index.php?title=Flatness_(manufacturing)&action=raw | Sir Joseph Whitworth popularized the first practical method of making accurate flat surfaces during the 1830s ... resulting in two surfaces which will closely agree each other, but which can still be concave or convex (and thus, not flat).
- CONFIRMED | the hand-scraping tradition and the 1840 paper's presentation to the British Association: the CircuitousRoot hand-scraping library | https://circuitousroot.com/artifice/machine-shop/surface-finishing/hand-scraping/index.html | This is Joseph Whitworth's 1840 paper presented to the British Association at their meeting in Glasgow. / Hand scraping for precision surfaces starts with Whitworth.
- CONFIRMED | Whitworth's later measuring machine, said to resolve a millionth of an inch | https://archive.org/download/miscellaneouspa00whitgoog/miscellaneouspa00whitgoog_djvu.txt (1856 address to the Institution of Mechanical Engineers, Glasgow) | a small machine, by which a difference in length of the one-millionth part of an inch is at once detected
- UNVERIFIABLE | Shop guides describe rotating the pairings rather than repeating one fixed sequence, and we report that as stated practice. | https://en.wikipedia.org/w/index.php?title=Flatness_(manufacturing)&action=raw ; https://ericweinhoffer.com/blog/2017/7/30/the-whitworth-three-plates-method ; Whitworth on Wikisource, https://en.wikisource.org/wiki/Miscellaneous_Papers_on_Mechanical_Subjects/A_Paper_on_Plane_Metallic_Surfaces_or_True_Planes | Wikipedia: The key insight is to then lap them against a third surface, C, with rotation. The various pairs of surfaces are lapped together in succession / Weinhoffer: The process can be completed in six steps, and then repeated until the desired level of flatness is achieved. / Whitworth: These parts compose an entire series, by repeating which a gradual approach is made to absolute truth; Rotating the pairings is supported. No guide found contrasts this with 'repeating one fixed sequence'; the guides checked, and Whitworth himself, describe a fixed series repeated. The page hedges it as reported practice. No finding.
- OBSERVED | this 1-D model converges under every order we tried | public/strata/flat-out-of-nothing/index.html:606-615 (the page's 1-D lapping model: start bows K0 = [8,-3,5], lapPair) ; recreated: checker A scratch/A/scripts/orders.mjs and checker B work/arith.mjs (all 6 fixed pairing orders over many random starts, plus random pair sequences) | A: 6 orders x 5 starts: worst final/start = 9.45e-5 / 200 random sequences of 300 laps: worst final/start = 7.98e-21. B: all 6 fixed orders, 12000 random starts, 60 cycles: worst final/start = 1.69e-27; random pair each step, 300 steps: worst = 1.34e-19
- CONFIRMED | Full-aperture absolute flatness is achievable. Rotations, N-position averaging, and shifted-map methods recover far more than one diameter. | https://wp.optics.arizona.edu/jcwyant/wp-content/uploads/sites/13/2016/08/Absolute-testing-of-flats.pdf (Ai & Wyant, Appl. Opt. 32, 4698, 1993) | the four measurements, M1, M4, M5, and M6, are necessary for solving the exact profiles along the diameter of every 45° angle ... Adding two more measurements, the profile in the area between two adjacent diameters of these diameters is also obtained with some approximation
- CONFIRMED | moved from microns to nanometres | https://opg.optica.org/ao/abstract.cfm?uri=ao-37-25-5951 (Parks, Shao, Evans 1998) | actual data from 150-mm-diameter surfaces that are flat to less than 25 nm peak to valley.
- CONFIRMED | a bootstrap hiding in Victorian shop practice | https://archive.org/download/miscellaneouspa00whitgoog/miscellaneouspa00whitgoog_djvu.txt (Preface) | The paper upon plane metallic surfaces was read at the British Association in Glasgow, in the year 1840; Whitworth's published practice of 1840 falls in Victoria's reign, which began in 1837; the Maudslay-era roots are earlier.
- CONFIRMED | Real plates flex under their own weight and the scraping load | https://en.wikisource.org/wiki/Miscellaneous_Papers_on_Mechanical_Subjects/A_Paper_on_Plane_Metallic_Surfaces_or_True_Planes | In getting up a surface of considerable extent, it is necessary to take into account the strain which the metal suffers from its own weight
- CONFIRMED | R = 1.56 km · 1/R = 6.40e-4 /m (5.0 µm bow over a 250 mm plate) | research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; recreated: node verify-flat-out-of-nothing.mjs (downloaded from /checks/ into an empty directory) and hand arithmetic 0.125^2/(2*5e-6) | ok default 5 um bow -> R = 1.56 km R=1562.5 m; ok bow 5 um -> R = 1.56 km, 1/R = 6.40e-4 /m, mate residual = 0
- CONFIRMED | contact 100% · gap RMS 0.00 µm | research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; recreated: node verify-flat-out-of-nothing.mjs | ok bow 5 um -> gap RMS 0.00 um over 100% contact
- CONFIRMED | The determinant of that little system is 2 (not zero) ... the unique solution a=b=c=0 | research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; recreated: node verify-flat-out-of-nothing.mjs; hand: det[[1,1,0],[0,1,1],[1,0,1]] = 1 + 1 = 2 | ok determinant of the three-mating system = 2 (nonzero -> unique solution) det=2
- CONFIRMED | a+b = 5.00 b+c = 2.00 c+a = 13.00; max bow 8.00 µm; Reset to 8, -3, 5 µm | public/strata/flat-out-of-nothing/index.html:606 (K0 = [8,-3,5]) and :221 (Reset button) ; recreated: hand arithmetic (checker B work/arith.mjs) | initial sums 5 2 13 max 8
- CONFIRMED | the alternating projection converges to the plane from any start | research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; recreated: node verify-flat-out-of-nothing.mjs; checker B work/arith.mjs | ok converges below 0.01% of the start bow in 9 cycles final 6.92e-4 um; B: worst final/start = 1.69e-27, strictly decreasing
- CONFIRMED | along the flip-axis diameter the three pairwise sums recover each absolute profile with exactly zero error (recovery error 0.00 nm) | research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; recreated: node verify-flat-out-of-nothing.mjs | ok three pairwise sums recover all three absolute profiles exactly (max error 0) max/recovered - true/ = 0 nm
- CONFIRMED | a shared odd function added to all three flats leaves every pairwise measurement bit-identical, which the verifier confirms to the last digit | research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; public/strata/flat-out-of-nothing/index.html (instrument 3 phantom) ; recreated: node verify-flat-out-of-nothing.mjs; checker B with the page's own phantom over amplitudes -40..40 | ok adding a shared odd bow leaves EVERY pairwise measurement bit-identical max/delta measurement/ = 0 nm
- MINOR | every number recomputed live and re-checked offline in research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs | research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs ; recreated: node verify-flat-out-of-nothing.mjs in an empty directory; grep of the verifier for import, require(, readFile, process.argv and fetch( (no hits) | exit 0, 27 ok, 0 FAIL, ALL CHECKS PASSED; aside: "it cannot notice a figure that drifted on this page"; F8. True of the mathematics and of the values as written; overstated as a check bound to this page.
- MINOR | Run it yourself: node research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs | https://artwaste.land/checks/research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs (published copy of research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs) ; recreated: curl into an empty directory, then node verify-flat-out-of-nothing.mjs | 200, 14887 bytes, byte-identical to the repo file; exit 0; ALL CHECKS PASSED; F9. It runs once fetched, but the instruction itself has no link; the download link is only in the aside.
- CONFIRMED | It is published, and it runs on its own: one file, and Node. | https://artwaste.land/checks/research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs (published copy of research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs) ; recreated: curl into an empty directory; node verify-flat-out-of-nothing.mjs (node v22.22.0) | exit=0; 27 ok; 0 FAIL; ALL CHECKS PASSED
- OBSERVED | Measured: with every file this layer ships emptied, it still passes. | research/verifier-independence/depends-1.json; research/verifier-independence/grading.json; recreated: grep of the verifier (no file reads) and a run in an empty directory | depends-1.json: {"rel": "research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs", "tier": "C0", "exitCode": 0, "depends": false}; grading.json: "opened nothing that /strata/flat-out-of-nothing/ ships"
- OBSERVED | Last verified 20 July 2026, when it was written. | git log -- public/strata/flat-out-of-nothing/index.html research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs: b03ce7088d 2026-07-20 09:25 +1000 (adds page and verifier) ; memory/log.d/2026-07-20-starved-veins-content-wave.md:86 | Every verifier passed and the headless gate passed on all six before the skeptics ran.; Still passes today (2026-10-11): exit 0, 27/0.
- CONFIRMED | relates notes: a perfect straight line drawn from bar-lengths alone, by turning a circle inside out; a diameter tolerance mistaken for a finish spec; Gauss's ant ... Egregium proves curvature is intrinsic | repo: public/strata/drawn-straight/index.html (href of eighty-years-to-a-straight-line), public/strata/smoother-than-a-billiard-ball/index.html, public/strata/egregium/index.html; https://en.wikipedia.org/w/index.php?title=Theorema_Egregium&action=raw | Peaucellier's 1864 cell draws an EXACT one, by turning a circle inside out / The slogan reads a diameter tolerance as a surface finish. / The theorem says that Gaussian curvature can be determined entirely by measuring angles, distances and their rates of change on a surface
- MINOR | README source record: 'Ma, S. et al., Pixel-based ...'; Grace's Guide 'supporting the precedence claim'; CircuitousRoot 'situates the practice in the precision-workshop tradition (Maudslay)' | https://api.crossref.org/works/10.1016/j.optlaseng.2017.10.021 ; https://www.gracesguide.co.uk/Joseph_Whitworth ; https://circuitousroot.com/artifice/machine-shop/surface-finishing/hand-scraping/index.html | Crossref authors: Dede Zhai, Shanyong Chen, Shuai Xue, Ziqiang Yin / Grace's: while here he made his first great discovery / CircuitousRoot: Hand scraping for precision surfaces starts with Whitworth. (0 hits for 'maudsl', 48 for 'whitworth'); F6.

## Findings
- [fixed] F1 (WRONG): The dek, meta, og, h2 and body said the laser "did not escape" (og: "never escaped") the bootstrap and that a Fizeau interferometer's reference flat is learned only by comparing flats pairwise, though a liquid surface levelled by gravity (PTB's mercury mirror, Germany's national primary flatness standard in 2002) is an absolute reference outside the flats; the h2 (id kept) and og now say "The laser still mostly leans on the bootstrap", the dek and meta say the usual calibration is again three flats compared pairwise (the dek adds that Schulz and Schwider "also name the other route: a liquid surface, levelled by gravity"), the body says an older absolute test compares the reference with a liquid surface levelled by gravity while the usual route is three flats measured pairwise and full-aperture three-flat tests have largely superseded the liquid mirror, the instrument tab reads "The bow flipped comparisons cannot see", and the verifier comments say the usual laser calibration still runs on the bootstrap.
- [fixed] F1 (WRONG): The details note said "every reference flat is still calibrated against other flats, and the bootstrap is closed, never escaped", though PTB's mercury mirror and PTB's angle-measuring method, which needs no external flatness standard, take their reference from outside the flats; it now says "most reference flats are still calibrated against other flats", that these methods close the bootstrap rather than escape it, and that the routes outside it take their reference from gravity, which levels a liquid surface, or the straight path of a light beam, and the footer adds Griesmann, Wang and Soons 2007 and PTB News 3/2002.
- [fixed] F2 (WRONG): The page cited Schulz and Schwider 1967 for the one-diameter limit and gave going beyond one diameter to "the rotation extensions (Fritz, 1984)", though Schulz and Schwider's own 1967 method reached arbitrary central sections from four interferograms; the page now ties the limit to three comparisons alone (the "classical three-comparison test", with Griesmann 2006 cited in the footer for it), says Schulz and Schwider's 1967 method "used a fourth interferogram to reach any diameter you choose" and that Fritz fit Zernike polynomials to the whole surface in 1984, carries the same correction in the dek, meta description, JSON-LD, HTML comment, md body, README and verifier comments, and no longer says the Zhai et al. survey states the one-diameter limitation, only that it cites both papers.
- [fixed] F3 (MINOR): The footer said Wikipedia's Surface plate article states that the method solves the bootstrapping problem of the first flat surface, a sentence that is in Flatness (manufacturing); the footer now cites Surface plate for surface plates taking their precision from no more-precise standard (and, citing Rolt 1965, for Maudslay's sets of three and the writers who credit Whitworth) and cites Flatness (manufacturing) for the no-other-reference and bootstrapping sentence.
- [fixed] F4 (MINOR): The footer said the Wikisource text "states neither venue nor date", though the chapter's footnotes date it to 1840 and the same edition's Preface gives both; the footer now says the chapter page carries the text "(its footnotes date it to 1840) but not the venue" and links the Wikisource Preface, quoting that the paper "was read at the British Association in Glasgow, in the year 1840".
- [fixed] F5 (MINOR): The title-variance note left out the British Association's own title for the paper; the body now says the Association's Report of the 1840 meeting abstracts it as "On producing True Planes or Surfaces on Metals", the footer links that Report on archive.org (p. 192) and says "a title as read there" for Grace's Guide, and the README lists all three variants.
- [fixed] F6 (MINOR): The layer's README (the source record, not the page) named the survey's author as "Ma, S.", said Grace's Guide supports Maudslay's precedence and said CircuitousRoot places the practice in Maudslay's tradition; the README now names Zhai, Chen, Xue and Yin with Optics and Lasers in Engineering 102, 92-99 (2018), says Grace's obituaries credit the discovery to Whitworth while he was at Maudslay's, drops the CircuitousRoot Maudslay clause and adds an entry for Nasmyth's 1883 autobiography.
- [fixed] F7 (MINOR): The page said flatly, with no source, that the practice already lived in Maudslay's workshop before Whitworth, though Whitworth's obituaries credit him with the discovery; the body now says Whitworth "probably did not discover it alone", that Maudslay's standard planes were made three at a time, rubbed each on each and finished with scrapers, that Nasmyth (who arrived there in 1829) called the art "a very old mechanical dodge" and that Whitworth's obituaries credit Whitworth himself, the Attribution note says the method "by most accounts predates Whitworth", the relates note adds "by most accounts", and the footer cites Nasmyth's autobiography (1883), ch. VIII, pp. 148-149 and Surface plate citing Rolt (1965).
- [fixed] F8 (MINOR): The dek said every number was "re-checked offline" by the verifier, and the check box and footer that the figures were reproduced or recomputed offline, though the verifier recomputes the mathematics and never reads the page; the dek now says "the mathematics recomputed independently offline ... (which checks the results, not this page's text)", the check box says the numbers are "recomputed independently offline, from the mathematics rather than from this page's text", and the footer says "the mathematics behind it independently offline".
- [fixed] F9 (MINOR): The check box's "Run it yourself" line gave a bare repository path with no link, which fails in an empty directory; it now says to download verify-flat-out-of-nothing.mjs (linked to /checks/research/flat-out-of-nothing/verify-flat-out-of-nothing.mjs) and run node verify-flat-out-of-nothing.mjs (Node only, no dependencies).