identity
This entry covers two operational identities that share one bar of metal or wood. They carry different hazards and different tolerances and are not interchangeable.
- Reference / comparator. A rigid bar, rule or narrow reference face whose working edge or face has a known or verified straightness tolerance. It compares a sampled line on a board, tool or machine surface with that working edge or face. A narrow working edge checks straightness along a line; a granite straightedge may instead have a long narrow working face. Keep edge straightness and face flatness explicit [SE1][SE10].
- Cutting or routing guide. A straight bar used as a fence for a knife, router or portable saw. This is a fixture inside the host tool's own procedure, not a measuring use, and a calibrated reference edge is not the default choice for it (see how NOT to use it).
- name: Straightedge.
- family / category: Measuring & layout → the reference-standard branch. Distinct from the Combination Square (which the corpus's own entry frames as a short-range straightedge substitute via its withdrawable blade: "not a long, unsupported straightedge... the withdrawn blade can sag over a wide span if left unsupported, reintroducing the error the tool exists to eliminate") [CS-cross], and from winding sticks (a paired straightedge-derived tool that reads twist, a different geometric property) [SE10].
-
terms kept apart:
straightnessis a line/form condition;flatnessapplies to a surface;squarenessis angular;level/plumbis gravity-referenced;twist/windis a surface relationship made visible by separated reference lines. A ruler may be straight enough for layout without being a calibrated or tolerance-specified reference straightedge. (Terminology per the review.) - one-line: A comparison reference with no moving parts and no user adjustment. A straightedge compares a sampled line on the work with a working edge or face of known or verified straightness. Its usefulness is bounded by its tolerance, calibration or verification status, length, support, orientation, condition, and the resolution of the gap-reading method. Precision examples still require verification or calibration.
- a genuinely humble, easily overlooked tool. Jim Tolpin, in a By Hand & Eye post dated March 14, 2016, builds his own wooden straightedge by one documented method (see variations) and tests it by the scribe-and-flip method [SE31]. (The original run attributed a "shoddy array of straight-ish, straight edges" quotation to Richard Maguire; no Maguire page was inspected in the correction pass, so that quotation is removed.)
✦ how NOT to use it
Rebuilt by mode (the reviewer, Fix 1). Passive comparison use and guide use must not share one safety label.
-
not-for:
- Checking 90° angles; straightness is not squareness [SE35][SE36].
- Checking plumb or level; a different, gravity-referenced property.
- Serving as the sole flatness reference without first verifying the straightedge itself.
- Asserting certified flatness from one placement or one favorable orientation.
- Matching machine-table coplanarity without the machine manual's datum and procedure.
-
dangerous — by mode. This class is no longer left "empty." The original text said that no straightedge-specific injury report was located. That is a research result about one search, not a risk assessment. Do not use the absence of straightedge-specific injury reports as proof that guide-mode operations are safe. (The reviewer)
Mode A: passive reference / comparator use (generally low hazard, not hazard-free).
- Heavy steel or granite tools can pinch fingers, damage the work, or injure feet if dropped; thin or knife edges and burrs can cut. (The reviewer's safety notes, relayed.)
- Inspect before use for burrs, dents, corrosion, cracks, bent sections, and damaged handles or lifting holes where present; clean the tool and the work before contact.
- Handle long or heavy precision straightedges with the method and number of people the maker specifies. Support long steel straightedges at their marked suspension points [SE14][SE42]. Protect the reference face in transport and storage.
- Keep fingers clear of the contact line when sliding or laying a heavy edge; do not press on a thin rule hard enough to bend it.
- General woodworking-injury statistics exist (one industry source cites roughly 720,000 annual woodworking injuries, hands and fingers most common) but none ties to a straightedge as the causal tool [SE37]; they are context, not a risk figure for this tool. Mode B: knife guide.
- Use a cut-resistant guide appropriate to the blade, keep the supporting hand out of the projected cut path, make controlled passes, and replace a nicked guide that can steer the blade. "You don't want your knife to hit a nick in the ruler while you're cutting, as it can suddenly steer your knife off its path" (trade blog; steel and hard-metal edges resist nicking better than plastic or wood) [SE19]. (Controls per the reviewer's safety notes, relayed; no knife-maker manual was consulted.)
-
A hold-down hand placed on the guide should sit behind the guide's edge, well clear of the blade line.
[manual, relayed: not independently sourced]Mode C: router guide. Bosch's GKF125 router manual (read through a fetch, "Guiding the Router"): "it is often necessary to guide the tool in a line parallel to a straight edge," and to get a straight cut, "securely clamp a board or other straightedge to the work surface, and guide the edge of the router sub-base along this path" [SE44]. Beyond that wording (manual text, read this pass): the guide is only a fixture inside the router's own procedure. - Securely clamp both the guide and the work; do not hold a guide by hand, since a handheld guide can slip into the cutter path.
- Verify the base-to-bit offset (the distance from the edge the base rides to the cutting edge of the bit) before the pass, and confirm clamp clearance so the base does not strike a clamp.
-
Follow the router manual's feed direction, guard, bit, dust and PPE instructions for the exact model.
[manual, relayed](the reviewer's safety notes; only the clamping sentence above was read). Mode D: portable circular or plunge saw guide. Bosch's GKT13-225 saw manual (read through a fetch): "a board or strip can be secured with clamp C (sold separately) to the workpiece as a straight edge guide"; "Ensure that the clamps do not interfere with free movement of the saw"; "support the panel or board close to the cut"; "Support large panels to minimize the risk of blade pinching and kickback"; supports go "under the panel on both sides, near the line of cut and near the edge of the panel" [SE45]. The blade "will tend to bind, causing KICK-BACK" when large panels are unsupported [SE45]. -
Secure the guide, support the sheet close to the cut, keep clamps clear of the saw's path, set depth correctly, keep the guard working, and control kickback per the exact saw manual.
[manual, relayed]for depth, guard and body-position points beyond the passages quoted above. -
OSHA states that powered handheld saws carry substantially greater hazards and require effective guarding
[manual, relayed][SE46]; use the exact host-tool manual for operating detail. Mode E: all powered-guide use. The guide inherits the host tool's full work-support, guard, bit/blade, feed, PPE and kickback requirements. Do not run a blade or router base against a precision reference edge unless its maker permits that use; prefer a straight sacrificial guide intended to be clamped. A reference-grade working edge that is nicked or worn is no longer a reference.
-
damaging:
- A nicked edge used as a cutting guide steers the blade (see Mode B) and ruins the edge as a reference [SE19].
- Dropping a precision-ground straightedge can introduce a bend or ding too small to see but large enough to corrupt readings. A Sawmill Creek poster says he would simply buy a new one rather than re-verify; practitioner statement, not a maker's guidance [SE7]. A dropped reference should be re-verified or replaced.
- Aluminum straightedges are more vulnerable to impact and bend damage than stainless steel — carried at lower, aggregate confidence (search-summarized across product listings, no single named source) [SE-gap].
-
beginner-trap:
- Trusting an unverified reference through a whole run of real work (the Huber account) [SE7].
- Using the light-gap test alone and stopping there. No visible gap does not prove precise flatness; the eye resolves only so much [SE9]. Use a local gap check or a proper method where the tolerance demands it.
- Using an untested level's edge, or tubing or extrusions, as a straightedge on trust alone [SE7].
- Treating a "flip" or a three-way match as calibration. They are shop checks (see lineage).
-
lineage-misuse:
- Confusing straightness, flatness and squareness [SE35][SE36]. A tool proven straight along its edge says nothing about whether a whole surface is flat, and nothing about angle.
- Treating a tape measure's blade as a straightedge [SE19][TM-cross].
- Treating winding sticks and a straightedge as interchangeable [SE10].
- Using a precision-ground reference straightedge as the default cutting guide. The reviewer states: do not recommend a calibrated metrology edge as the default cutting guide; running a blade or router base against it can nick or wear the surface its accuracy depends on. The original run flagged this only as inference from product-category separation (commercial guide products such as the Woodpeckers Clamp-N-Guide sold separately from precision reference lines) [SE-gap].
correct use
- what it is genuinely good at: comparing a selected line on a board face or edge, a machine table, a fence, a bench top or another tool's claimed-flat surface against a working edge of known or verified straightness, and screening for flatness by repeating the comparison along several specified paths.
-
what a single placement samples (the reviewer, Fix 5). A straightedge checks straightness along a selected line and can screen a surface for flatness by checking multiple specified paths. It does not measure the entire surface simultaneously. One placement compares one line or contact band; a defect between sampled paths can be missed, and rocking on high points can change the apparent gap. For board faces, sample several lengthwise, crosswise and diagonal paths and pair with winding sticks for twist. In high-consequence matches, record the sampling directions and an acceptance tolerance. Certified flatness requires a surface-plate, indicator or optical method; Starrett's Planekator is a system with a calibrated granite straightedge, supports, indicator, carriage and correction data, far more than laying an edge down once
[manual, relayed]. -
reading the gap — two shop methods, both qualitative or local.
- Visual light-gap check. Lay the clean straightedge on the clean surface and sight against a light source. "Seeing no light is an automatic pass; seeing light only identifies a need for proper measurement" [SE9]. The same source says deviations of "1 or 2 microns from straightness can be detected by the eye with some practice," so an untrained eye will miss small gaps [SE9]. A gap that moves as the straightedge is repositioned points at the work; a gap that stays with the tool as it is turned points at the tool.
- Feeler gauge in the gap: a local line-gap check. A feeler gauge measures a local gap relative to the straightedge and its contact points, not automatically the minimum-zone flatness of the surface. Conditions: clean, undamaged surfaces; a straightedge of known tolerance and adequate length; multiple prescribed orientations; no forcing or pressure that changes contact; record the largest local gap; compare to a job-specific tolerance. Dirt, burrs, edge thickness, force, rocking, inaccessible interior gaps, line selection and the straightedge's own form error all matter. That Lee Valley sells feeler gauges near its straightedges shows a product pairing, not a validated procedure.
- measurement conditions. Cleanliness and burr control are part of the measurement. Do not press hard enough to bend a thin rule, seat it over dirt, or force a feeler gauge. Gravity, support, orientation, temperature, wear, contact pressure, viewing angle and the observer all affect the result (the reviewer; and NBS/Starrett support points above).
- verify the reference before trusting it. For high-consequence setup work, check the straightedge itself: scribe-and-flip for a wooden or shop edge (a check, not calibration); calibration or verification by a metrology lab, or against a surface plate, for a precision edge. NBS lists the straightness of straight edges among reference surfaces "calibrated by optical and mechanical techniques" (SP 250, read directly this pass) [SE43]. No user adjustment exists, but that is not the same as needing no calibration.
✦ matching — for the app's read
Matching splits into modes. Do not apply the low-hazard label of the reference mode to powered guide work.
-
mode field (split per the reviewer):
reference comparator·layout/scribing edge·hand-knife guide·router guide·portable-saw guide·screed/leveling beam·specialized notched reference. -
required reference fields:
working edge or face,length,stated tolerance/grade,calibration or verification status,support/orientation,material,thermal/environment condition,sampling path,gap-reading method,acceptance threshold. -
required guide fields:
host tool/model,guide material/profile,offset,clamp method,work support,feed/cut direction,clamp clearance,guard/PPE condition,sacrificial/reference status. - does (reference mode): checks straightness of a jointed or sawn edge along selected lines; screens board faces, machine tables and fences for flatness along multiple paths; supports machine setup checks (jointer, table saw, planer), where it is a comparison aid only. Match machine-table coplanarity only with the machine manual's datum and adjustment procedure; a straightedge alone does not define which surface or knife is the reference [SE32][SE33].
- does (layout mode): scribes and marks long straight lines.
- does (guide mode): acts only as a fixture inside a named host-tool procedure. See how NOT to use it for the conditions. Prefer a straight, clampable, sacrificial guide, not a calibrated reference edge.
- suits: workbench-top flattening (several paths, with winding sticks for twist and a feeler gauge for local gaps); table saw, jointer and planer setup checks; machinist checks of a surface plate or machine table (the precision branch).
-
kit-fit:
- pairs with: winding sticks for twist: "Straightedges do not indicate twist... whereas winding sticks are a pair of narrow boards with parallel edges that allow you to determine if a board, joint, or assembly is—or is not—free of twist" [SE10]. Feeler gauges for a local gap estimate, never labeled as complete surface flatness without a valid procedure. A combination square for shorter checks [CS-cross].
- substitutes for: a shop-made wooden straightedge for woodworking-grade layout (see variations) [SE31]. Removed pending a source and verification condition: the "known-flat door" substitution. A laser level is not offered as an interchangeable substitute: it would need the task defined first (straightness, alignment, level/plumb, range, required uncertainty).
- cannot replace: winding sticks for twist [SE10]; a spirit level for plumb/level; a square for 90°. "Straightness is concerned with the linearity of an object, while flatness is concerned with the planarity of a surface... straightness only makes sure that a single line on a surface has to be within the limits, flatness does the same for a collection of lines," and "squareness guarantees corners and edges meet at right angles" [SE35][SE36].
- not-for (project-keyed): an unverified or uninspected straightedge as the sole flatness reference for real work (the Huber account is a practitioner narrative of that failure, lower-tier evidence) [SE7]; a tape measure's case edge as a straightedge (the corpus's own Tape Measure entry names this as a commonly held trade assumption and flags it as unfilled) [TM-cross]; a 90° check; an untested level's edge or extruded tubing, since a working machinist in the thread warns "I wouldn't vouch for tubing to be straight, or any type of extrusions, either... They are usually curved or twisted to some degree" [SE7].
- access (reference mode only): passive comparison is generally low hazard, not hazard-free, and has a low skill floor to use but a real learning curve to interpret (see the light-gap notes). It is not "genuinely beginner-safe" in general: that label does not transfer to knife, router or saw guide use, and heavy steel or granite tools carry handling hazards (see how NOT to use it). The habit that separates a safe beginner from an exposed one is verifying the reference before trusting it.
- revival: a case of a tool whose value is the discipline it supports rather than any capability of its own: it has no technique to master, and the payoff is catching a silent error (a warped bench, an out-of-parallel fence, an unverified reference tool) before it propagates into finished work. Almost any shop can buy or make one; the work is in knowing what its tolerance and condition let it say.
- Treat current prices and availability as volatile metadata, not durable matching attributes.
anatomy
- the working edge or face — the ground, lapped, scraped or hand-planed reference surface itself. Precision steel may be a single edge (380 series, unbeveled) or beveled/graduated (385/387) [SE14]; a granite straightedge may have a narrow working face finished flat.
- bevel — some steel straightedges bevel one edge to a thin profile for light-gap sensitivity; "knife-edge straightedges" are a dedicated metrology subtype [SE9][SE14].
- graduated scale (optional) — Starrett 387 series and Woodpeckers SERX carry an engraved scale, a secondary scribing function [SE14][SE30].
- suspension/support points — Starrett's page: "The 380 and 385 Straight Edges in sizes 36" and longer are marked with arrows at two suspension points," to be suspended at those two points when used on edge "to minimize deflection" (read directly this pass) [SE14][SE42]. Long edges deflect: support points, horizontal or vertical orientation, length, section stiffness, and whether the tool is carried to the work or the work to the reference all belong in the setup; gravity changes the result.
- handling features — Heavy steel or granite tools may carry lifting holes or handles; inspect them (see how NOT to use it). Beyond these, no further named anatomical parts were found: a searched absence.
variations (living)
Each variation has its own tolerance and permitted-use fields; do not carry one branch's numbers to another.
- wooden straightedge / batten — Jim Tolpin's method (one shop-made method). Per Tolpin (read directly): radial-grain stock, usually Honduras mahogany or cherry; cut to length and resawn to about 3/8 inch; stood on edge for several weeks in the shop's own environment to cull warped or bowed pieces; top edge tapered or curved and thickness tapered from bottom to top; hand-rubbed tung oil, then several coats of shellac as a vapor barrier; tested by scribe-and-flip, with occasional retests [SE31]. Evidentiary level: his moisture-equilibration explanation (more end grain exposed, faster adjustment) is explicitly theoretical ("In theory anyway"). He calls the stock "rift-sawn (sometimes called radial cut)"; per the reviewer, a commenter challenged that terminology versus radial/vertical grain, so this entry uses radial/vertical grain. Species, grain, conditioning time, section and finish interact with the shop environment and the required tolerance; the post is not a controlled stability study. Any wooden reference needs periodic re-verification. Permitted use: layout and board checks within a few thousandths at best (see scribe-and-flip limits above); not a calibrated reference.
- precision-ground steel straightedge — Starrett 380 (unbeveled, ungraduated, recorded in the original run at ±.0002"/ft), 385 (beveled), 387 (beveled and graduated); Kinex DIN 874-graded lines [SE12][SE13][SE14]. DIN grade values: see history; none quoted here.
- granite straightedge — Starrett product grades (not a universal granite rule): laboratory grade AA .000025 in per 12 in and inspection grade A .000050 in per 12 in, per the Starrett granite page for the straightedge section (read directly this pass; the original entry wrongly gave "per 6 in", the denominator for other granite accessories) [SE15]. Other makers: Precision Granite (US), Eley Metrology (UK) [SE16]. Do not mix surface-plate, master-square, angle-plate, parallel and straightedge tolerance tables.
-
knife-edge straightedge — one working edge beveled and very slightly rounded, for light-gap testing; DIN 874 treats it as its own part
[manual, relayed][SE9]. - aluminum box-beam / I-beam straightedge (trade) — flooring, tile, drywall, concrete; T-5/T-6 aluminum, 2'-8' lengths; screeding and leveling, not a reference-grade tool [SE17][SE18].
- woodworking-marketed aluminum/steel straightedge — Veritas (Lee Valley) and Woodpeckers SERX, with a self-standing inverted-vee-groove profile; current market entries in proof of existence [SE32][SE33][SE30].
-
luthier variant — StewMac's Precision Notched Straightedge. Per the reviewer, a notched luthier straightedge reads the fingerboard or neck between frets for a specified scale pattern; it does not "read fret relief" through the fret crowns and is not a universal guitar straightedge
[manual, relayed][SE34].
trades
- Machinist / toolmaking — the precision branch; DIN 874 grading and Starrett's precision lines are machine-shop metrology [SE1][SE12][SE14].
- Cabinetmaking / fine furniture — hand-conditioned wooden battens and the modern woodworking-marketed steel and aluminum lines [SE31][SE32][SE30].
- Flooring / tile / drywall / concrete — the aluminum box-beam / I-beam branch, for screeding and leveling [SE17][SE18].
- Luthiery — the notched variant, with its limits noted under variations [SE34].
- Precision manufacturing / CNC and CMM calibration — the granite branch [SE15][SE16].
lineage
Four branches, not one origin (the reviewer, Fix 3). The tool class reaches back well past precision metrology. The history below is kept in four branches, and the Maudslay and Whitworth milestones are shown only in the branch they belong to.
-
1. Straight rulers, drawing edges and craft battens. A dated reference to a straight rule predates the nineteenth century: The reviewer cites the title of Joseph Moxon's Mechanick Exercises (Folger record, 1703 third edition), which refers to work "with the help of a straight Ruler and a pair of Compasses"
[manual, relayed][SE38]. This shows a straight ruler in use in 1703; it does not show a reference-grade straightedge or a woodworking batten. - 2. Woodworking reference straightedges. The Smithsonian National Museum of American History holds an 18-inch mahogany straightedge, catalog number MC.260301.004, in its Work and Industry: Hand Tools collection; the record gives no date or place of manufacture [SE39]. Use it as proof that the object type exists in a museum, not as a priority date. The modern wooden-batten practice documented here is Tolpin's (see variations).
-
3. Precision metal and granite reference surfaces. This is where Maudslay and Whitworth belong. Henry Maudslay kept a bench micrometer nicknamed "the Lord Chancellor" in his London workshop from about 1805, described by the Science Museum as the shop's "court of appeal"
[manual, relayed][SE2][SE3]. It is a micrometer, not a straightedge. Joseph Whitworth's three-plate method (scraping and Prussian-blue transfer between three trial plates worked in pairs; described to the British Association in 1840 and presented to the Institution of Mechanical Engineers in 1856) is important to generating precision plane surfaces[manual, relayed: IMechE][SE1][SE4][SE5]. Neither is established here as the direct ancestor of every modern ground straightedge. Modern descendants: Starrett 380/385/387 steel straightedges and Kinex DIN 874-graded lines [SE12][SE13][SE14]; granite straightedges, non-magnetic and thermally stable, used for CNC/CMM and surface-plate work [SE15][SE16]. - 4. Trade screeding and cutting guides. Aluminum box-beam / I-beam straightedges used in flooring, tile, drywall and concrete work (screeding, floor leveling), a lower-precision, higher-span-strength branch [SE17][SE18]; and clamped boards or strips used as saw and router fences. The corpus's own Track Saw entry documents the shop-made clamped-strip technique as the track saw's direct ancestor [TS-cross].
-
contemporaries — the corpus's own siblings that get mistaken for it:
- Winding Sticks — a paired, related tool built to read a different property: "Straightedges do not indicate twist... winding sticks are a pair of narrow boards with parallel edges that allow you to determine if a board, joint, or assembly is—or is not—free of twist," and "a single winding stick can double as a straightedge" (one direction only) [SE10].
- Combination Square — its own entry frames its withdrawn blade as a short-range straightedge stand-in with a stated sag limitation [CS-cross].
- Sliding Bevel — the corpus's own entry states: "A bevel is not a straightedge. The blade is thin spring steel, slotted, and not ground for straightness; ruling long lines against it certifies its flex" (labeled there as Haft's own structural inference) [SB-cross].
- A tape measure's blade — deliberately curved (crowned) to self-support over a span; the same geometry that makes it work as a tape measure is what makes it fail as a flat reference [SE19].
Three related comparison ideas, kept separate (the reviewer, Fix 4)
| Method | What it is | Purpose and limit |
|---|---|---|
| Scribe-and-flip | Draw a line along the edge, reverse the tool, draw a second line over the first; two lines mean disagreement between the two orientations [SE7][SE31]. | A shop verification check, not calibration to a stated tolerance. The result depends on how the edge is reversed (face-to-face, end-for-end, or about the reference line each expose different error components), line width, registration and viewing resolution. Jamie Buxton (forum) describes it as good to roughly five thousandths for woodworking [SE7]. |
| Comparing three straightedges | A metrology reference site says: "Three straight edges can be used to test and calibrate themselves to a certain extent... two edges might match if one is a mirror of the other but three edges which match each other must be straight" [SE9]. | Lower-tier source, and its own words are "to a certain extent." It does not control twist. No defined set of pairings and orientations was in hand, so this entry does not recommend it as a procedure; a real procedure and source are needed first. |
| Whitworth's three-plate method | Iterative marking and scraping of plane surfaces in paired combinations to generate true planes [manual, relayed: IMechE] [SE1][SE4][SE5]. | Belongs to precision-plane manufacture. The modern NIST three-flat calibration is a defined absolute method that reports a profile along specified lines, not a generic "three things agree" rule [manual, relayed]. |
Inference, labeled. The conceptual resemblance among the three (comparing a thing against others to expose its own error) is the reviewer/Haft inference, not direct procedural continuity. The earlier Petroski "form follows failure" fit remains Haft's own structural reading, with no source tying Petroski to straightedges [SE6].
history
-
No single origin. See the four branches in lineage. The earliest dated item located in the corrected entry is the 1703 Mechanick Exercises reference to a straight ruler
[manual, relayed]; the Smithsonian straightedge is undated. -
Precision branch milestones. Maudslay's c.1805 micrometer (Science Museum Group record
[manual, relayed]) and Whitworth's 1840 and 1856 papers on true planes are milestones in the precision-engineering branch only [SE1][SE3][SE4][SE5]. Adjacent objects (Whitworth's surface plates and 1855 measuring engines in the Science Museum [SE1]; Maudslay's micrometer) are not proof of a straightedge object. -
Company dates, not category dates. Kinex (Czech Republic) is recorded as founded 1906 [SE13]. Starrett was founded in 1880 and has been privately owned by a MiddleGround affiliate since 2024; it currently markets straightedges
[manual, relayed][SE23][SE41]. No dated catalog establishing that the 380/385/387 line existed continuously from 1880 was located, so no product-line continuity is claimed. - Two traditions persist side by side. George Wilson, a self-identified machinist on a 2009 Sawmill Creek thread: "Being also a machinist, I have my own little 'Bureau of Standards.' These include 3 cast iron straight edges that are hand scraped to within .0001"... .0001" is of course not needed for woodwork!!!" [SE7]. Woodworking batten practice and machinist-grade practice serve different tolerance worlds.
-
Standards. DIN 874 has separate parts for flat steel straightedges and for knife-edge straightedges; the reviewer reports the DIN Media record for Part 2 (2003-11) identifies knife-edge straightedges and straightness tolerances
[manual, relayed]. Tolerances depend on part, tool form, nominal length and grade. When citing a DIN grade, recordstandard,edition,part,tool form,grade,nominal length,working face/edgeandmaximum permissible error, and link the official DIN record and the maker's certificate for the exact product. No DIN tolerance values are quoted here: the original run drew them from a reseller explainer [SE12] and they could not be tied to a named table. - Patents. In the sources checked, no foundational patent for a straightedge or its grinding/lapping process was identified. This is a search report, not evidence that none exists. Later, peripheral patents found: US4622755 (drafting-tool "parallel straightedge"), US5481810 (1996 combination tape measure and straight edge), US20190025038 (2019 straightness-checking-method application) [SE20][SE21][SE22]. No Starrett straightedge-specific patent number was confirmed [SE23].
✦ human side
- Joseph Whitworth and Henry Maudslay — the named figures of the precision-surface branch (not the whole tool class); see lineage branch 3 [SE1][SE2][SE3][SE4][SE5].
- L.S. Starrett Company. Laroy Sunderland Starrett founded the company in 1880 in Athol, Massachusetts, after patenting a combination square in 1878 [SE23][SE24]. The company currently markets the 380/385/387 straightedge series and a precision granite line [SE14][SE15]. This is company history and a current catalog, not proof of a continuous product line.
- Richard King and living hand-scraping craft culture. King runs week-long scraping courses ("King-Way") where students learn to scrape and inspect machine ways, often scraping their own straightedges and surface plates [SE25][SE26][SE27]. James Anderton, writing for Engineering.com, frames hand scraping as "the last bastion for the blacksmith's art in a CNC world" [SE28]. "Equaling and averaging" scraping is traced to Whitworth's disciples and to C.O. Griffin's 1898 American Machinist article [SE29].
- a named, dated practitioner account of the reference tool being the hidden fault (lower-tier evidence). In a 20-post Sawmill Creek thread, "Is your straight edge straight??," started by Bill Huber on October 25, 2009, Huber found his own straightedge gave inconsistent readings when flipped end-for-end, after real setup work (table saw wings, jointer, router table) had been done against it; a downstream symptom (crosscuts drifting out of square on longer boards) exposed it. Huber: "I don't really know how to test a straight edge unless you take it to a machine shop" [SE7]. Also in the thread (forum evidence, one poster's account each): glenn bradley would buy a new Lee Valley straightedge if his were dropped rather than trust a re-check; george wilson owns three hand-scraped cast-iron straightedges and a 30" granite laboratory-grade straightedge, recommends a mid-grade granite surface plate to verify a straightedge, and describes a taut fine-wire technique from ship-propeller-shaft alignment; harry strasil describes a .035" MIG-wire method with keystock risers and feeler gauges; Jamie Buxton describes scribe-and-flip as good to about five thousandths; scott vroom admits using an untested 24" level's edge on trust [SE7]. This preserves practitioner texture; it is not a standard.
- Jim Tolpin's wooden straightedge (one documented method, not a standard). By Hand & Eye Online, "Let's Get This Straight," posted March 14, 2016 by Jim Tolpin [SE31]. See variations.
- Removed in this pass: the Maguire / English Woodworker material. The original entry attributed quotations ("easy enough to make one accurately straight off a hand plane", a three-reasons list, an exchange with commenters Simon, Donnie Clapp and DaveM, and a "shoddy array" line) to Richard Maguire's The English Woodworker. The By Hand & Eye page does not contain that language, and no direct Maguire URL was supplied or inspected. The material will return only with a direct, inspected source. The practitioner disagreement it illustrated (shop-made versus purchased) is not asserted here.
✦ proof of existence
-
older and broader than the precision branch. A straight ruler appears in the 1703 Mechanick Exercises title
[manual, relayed][SE38]; an 18-inch mahogany straightedge is held by the Smithsonian (MC.260301.004, undated) [SE39]. Neither is a priority date for the precision branch. -
precision-branch milestones, as above: Maudslay's c.1805 micrometer
[manual, relayed]; Whitworth's 1840 and 1856 papers on true planes [SE1][SE3][SE4][SE5]. -
calibration of straight edges is a documented practice. NBS Special Publication 250 (1980) lists "the straightness of straight edges" among surfaces "calibrated by optical and mechanical techniques" [SE43]. The reviewer also cites NBSIR 85-3182 as showing calibrated straightedge shape and gravitational-sag correction
[manual, relayed][SE47]. - patents: no foundational patent identified in the sources checked (search report); see history.
-
company existence, not product-line continuity. Starrett, founded 1880 and privately owned by a MiddleGround affiliate since 2024, currently markets straightedges
[manual, relayed][SE23][SE41]. -
current market data — access-dated, volatile, excluded from historical confidence. Recorded by the original run as fetched in July 2026, not re-fetched in this pass. The reviewer found that the Starrett 380-24-with-certification page now shows a different price or status than the original table, so no Starrett price is repeated here. Treat everything below as a dated snapshot of offers, which proves only a dated offer and nothing about lineage:
- Veritas Steel Straightedge (Lee Valley, made in Canada): 12" $54.50, 24" $78.50, 36" $129.00 (July 2026, as recorded) [SE32].
- Veritas Aluminum Straightedge (Lee Valley): 12" $22.50, 18" $31.50, 24" $39.50, 38" $56.50 (July 2026, as recorded) [SE32].
- Woodpeckers SERX Straight Edge Rule (aircraft-grade aluminum, self-standing inverted-vee profile, recorded as 0.0015"/ft): 24" $79.99 sale, regular $89.99 (July 2026, as recorded) [SE30].
- Starrett 380 series: specification (steel, unbeveled, ungraduated, ±.0002"/ft as recorded) and SKU/UPC/NSN 5210-00-735-7604 were recorded in July 2026 from the Starrett checkout page; current prices, discounts and the 18" discontinuation status are volatile and the reviewer reports the table no longer matches [SE14].
- Kinex DIN 874/1-graded straightedges are sold through US distributors (Taylor Toolworks, Northwest Passage Tools) as recorded in July 2026 [SE13].
references
Tags SE8 is unused; tags SE38 to SE47 were added in the fix pass. [manual, relayed] marks a source whose content is carried from the review.
| Tag | Source | Kind |
|---|---|---|
| SE1 | Linda Hall Library, "Joseph Whitworth," Scientist of the Day, 2020-12-21, https://www.lindahall.org/about/news/scientist-of-the-day/joseph-whitworth/ | museum/library reference (carried from original) |
| SE2 | "Henry Maudslay," English and American Tool Builders (1916), via todayinsci.com, https://todayinsci.com/M/Maudslay_Henry/MaudslayHenry-ToolBuilders(1916).htm | book excerpt, historical (carried) |
| SE3 | Science Museum Group Collection, "Maudslay's Lord Chancellor bench micrometer," https://collection.sciencemuseumgroup.org.uk/objects/co59282/maudslay-s-lord-chancellor-bench-micrometer — supports c.1805, maker, object type, "court of appeal"; it does not make the micrometer a straightedge | museum collection record (relayed from the reviewer) |
| SE4 | EverybodyWiki, "Whitworth's three plates method," https://en.everybodywiki.com/Whitworth's_three_plates_method | encyclopedia (carried; low tier) |
| SE5 | Eric Weinhoffer, "The Whitworth Three Plates Method," 2017-07-30, https://ericweinhoffer.com/blog/2017/7/30/the-whitworth-three-plates-method | blog, technical (carried) |
| SE6 | Medium, summary of Henry Petroski, The Evolution of Useful Things, https://medium.com/fall-2014-nmix-4310-book-reports/the-evolution-of-useful-things-fd1b9891a365 | book summary (carried) |
| SE7 | Sawmill Creek Woodworking Community, "Is your straight edge straight??," started by Bill Huber, 2009-10-25, https://sawmillcreek.org/threads/is-your-straight-edge-straight.123208/ | forum, lower tier (carried) |
| SE9 | what-when-how.com, "Straight Edges (Metrology)," https://what-when-how.com/metrology/straight-edges-metrology/ | technical reference, lower tier (carried) |
| SE10 | Wikipedia, "Straightedge," https://en.wikipedia.org/wiki/Straightedge, and "Winding stick" | encyclopedia (carried) |
| SE11 | Removed. Formerly a merged Maguire / By Hand & Eye citation; split in the fix pass (see SE31) | (withdrawn) |
| SE12 | Wonkee Donkee Tools, "What accuracy and tolerance are engineer's straight edges made to?," https://www.wonkeedonkeetools.co.uk/engineers-straight-edges/what-accuracy-and-tolerance-are-engineer-s-straight-edges-made-to | trade reference; no values used (carried) |
| SE13 | Taylor Toolworks, "Kinex DIN 874/1 Straight Edges," https://taytools.com/products/kinex-din-874-1-straight-edges | product (carried) |
| SE14 | Starrett, "380-18 Steel Straight Edge," https://checkout.starrett.com/380-18-steel-straight-edge/ (fetched July 2026 by the original run; not re-fetched) | product (carried; volatile data excluded) |
| SE15 | Starrett, "What You Need to Know about Precision Granite Inspection Products," https://www.starrett.com/news-events/what-you-need-to-know-about-precision-granite-inspection-products — straightedge grades AA .000025 in per 12 in, A .000050 in per 12 in | manufacturer technical content (read in fix pass) |
| SE16 | Eley Metrology, "Granite Metrology Standards," https://eleymet.com/metrology-equipment/metrology-standards/ | manufacturer (carried) |
| SE17 | Barwalt Straight Edges, TileTools, https://tiletools.com/products/barwalt-straight-edges | product (carried) |
| SE18 | Bon Tool Aluminum Straight Edge Set, Home Depot, https://www.homedepot.com/p/Bon-Tool-Aluminum-Straight-Edge-Set-14-302/305926774 | product (carried) |
| SE19 | Tiger Fish Tools, "How To Cut A Straight Line With A Utility Knife (7 Tips!)," https://www.tigerfishtools.com/news/utility-knives-and-cutting-techniques/straight-line-utility-knife/ | trade blog (carried) |
| SE20 | US Patent 4622755, "Parallel straightedge," https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/4622755 | patent (carried) |
| SE21 | US Patent 5481810, "Combination tape measure and straight edge apparatus" (1996), https://patents.google.com/patent/US5481810 | patent (carried) |
| SE22 | US Patent Application 20190025038, "Straightness checking method," https://patents.justia.com/patent/20190025038 | patent application (carried) |
| SE23 | Wikipedia, "L. S. Starrett Company," https://en.wikipedia.org/wiki/L._S._Starrett_Company | encyclopedia (carried) |
| SE24 | FundingUniverse, "History of L.S. Starrett Company," https://www.fundinguniverse.com/company-histories/l-s-starrett-company-history/ | company history (carried) |
| SE25 | The Hobby-Machinist, "[Noteworthy] Scraping In A Precision Straight Edge," https://www.hobby-machinist.com/threads/scraping-in-a-precision-straight-edge.34485/ | forum (carried) |
| SE26 | The Hobby-Machinist, "Thoughts on Richard King scraping and machine reconditioning class SLC April 2022," https://www.hobby-machinist.com/threads/thoughts-on-richard-king-scraping-and-machine-reconditioning-class-slc-april-2022.99241/ | forum (carried) |
| SE27 | The Hobby-Machinist, "King-Way Scraping Class," https://www.hobby-machinist.com/threads/king-way-scraping-class.108598/ | forum (carried) |
| SE28 | James Anderton, "Hand Scraping: the Last Bastion for the Blacksmith's Art in a CNC World," Engineering.com, 2015-03-12 (updated 2024-05-01), https://www.engineering.com/hand-scraping-the-last-bastion-for-the-blacksmiths-art-in-a-cnc-world/ | trade publication (carried) |
| SE29 | Circuitous Root, "Hand Scraping (For Precision Surfaces)," https://circuitousroot.com/artifice/machine-shop/surface-finishing/hand-scraping/index.html | technical reference (carried) |
| SE30 | Woodpeckers, "SERX Straight Edge Rule," woodpeck.com (fetched July 2026 by the original run) | product (carried; volatile) |
| SE31 | Jim Tolpin, By Hand & Eye Online, "Let's Get This Straight," posted 2016-03-14, https://www.byhandandeye.com/lets-get-this-straight/ — one practitioner's method; moisture explanation explicitly theoretical | blog, practitioner account (read in fix pass) |
| SE32 | Lee Valley, "Veritas Steel Straightedge" and "Veritas Aluminum Straightedge," leevalley.com (fetched July 2026 by the original run) | product (carried; volatile) |
| SE33 | Woodpeckers product marketing copy (table saw/jointer/router table alignment use case), woodpeck.com | product marketing (carried) |
| SE34 | StewMac, "Precision Notched Straightedge," https://www.stewmac.com/luthier-tools-and-supplies/types-of-tools/straightedges/precision-notched-straightedge/ | product (carried) |
| SE35 | thecontentauthority.com, "Straightness vs Flatness: Which One Is The Correct One?," https://thecontentauthority.com/blog/straightness-vs-flatness | technical reference, lower tier (carried) |
| SE36 | WoodnBits, "What 'Flat, Square, and Straight' Really Means in a Real Shop," https://woodnbits.com/woodworking-basics-safety/flat-square-straight-meaning/ | trade blog (carried) |
| SE37 | WWGOA, "Shop Accident Statistics & Woodworking Safety," https://www.wwgoa.com/post/shop-accident-statistics-woodworking-safety | trade safety reference (carried) |
| SE38 | Folger Shakespeare Library, Moxon, Mechanick Exercises (1703), https://catalog.folger.edu/record/750916 — title refers to a "straight Ruler" | library catalog record ([manual, relayed]: page did not load in the fix pass; from the reviewer) |
| SE39 | Smithsonian National Museum of American History, Straightedge MC.260301.004 (catalog 260301.004), https://americanhistory.si.edu/collections/object/nmah_655343 — wood, mahogany, 18 in, no date listed | museum record (read in fix pass) |
| SE41 | MiddleGround / Starrett transaction completion, GlobeNewswire, 2024-06-05, https://www.globenewswire.com/news-release/2024/06/05/2894132/0/en/MiddleGround-Completes-the-L-S-Starrett-Company-Transaction.html — 1880 founding and 2024 private ownership; says nothing of a specific straightedge series | press release ([manual, relayed]) |
| SE42 | Starrett, Straight Edges, https://www.starrett.com/products/precision-measuring-tools/precision-hand-tools/precision-rules--straight-edges-and-parallels/straight-edges — 380/385 in sizes 36" and longer marked with arrows at two suspension points | manufacturer page (read in fix pass) |
| SE43 | NBS Special Publication 250, Calibration and Related Measurement Services (1980), https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nbsspecialpublication250e1980.pdf — Chapter II, Section D, item 2: "the straightness of straight edges are calibrated by optical and mechanical techniques" | government publication (read in fix pass) |
| SE44 | Bosch GKF125 router manual, "Guiding the Router," https://ocsmedia.boschtools.com/binary/ocsmedia/optimized/full/o200941v2_2610052572_0918_GKF125CE.pdf | manufacturer manual (passage read in fix pass via fetch summary) |
| SE45 | Bosch GKT13-225 saw manual, https://www.boschtools.com/us/en/ocsmedia/2610051844_0918_GKT13225.pdf — straight edge guide, clamp clearance, panel support, kickback (pp. 5, 13, 22) | manufacturer manual (passages read in fix pass via fetch summary) |
| SE46 | OSHA, Handheld Saws, https://www.osha.gov/etools/machine-guarding/saws/handheld | government safety authority ([manual, relayed]) |
| SE47 | NBS, Validation Metrology of the Large Optics Diamond Turning Machine, https://nvlpubs.nist.gov/nistpubs/Legacy/IR/nbsir85-3182.pdf | government report ([manual, relayed]) |
| SE-gap | Claims searched for directly but not sourced: aluminum-vs-steel impact vulnerability; precision-straightedge-as-default-cutting-guide caution (now also stated by the reviewer); flagged inline | honest gap marker |
| CS-cross | Combination Square (this corpus), withdrawn blade as a short-range straightedge substitute with a stated sag limitation | corpus cross-reference |
| SB-cross | Sliding Bevel (this corpus), distinguishes itself from a straightedge | corpus cross-reference |
| TS-cross | Track Saw (this corpus), clamped straightedge strip as its shop-made ancestor | corpus cross-reference |
| TM-cross | Tape Measure (this corpus), straightedge-substitution caution flagged there as an unfilled gap | corpus cross-reference |
Corpus cross-references are for consistency only and are not independent evidence for any mechanical claim. The original JN-cross to Jointer Plane supported the Maguire/Simon rule of thumb and is dropped with it.
The reviewer's corrections (12)
-
BLOCKER— 1. Split passive reference use from knife/router/saw guide use — Exact claim: The straightedge is “genuinely beginner-safe and low-skill-floor,” has “no motor, no blade, purely visual/tactile,” and the DANGEROUS class is “honestly and fully empty,” while matching also recommends it as a knife, router, and circular-saw guide. -
BLOCKER— 2. Replace “truth” and “nothing to calibrate” with comparator language — Exact claim: The tool has “no function but to tell the truth,” has “nothing to calibrate,” and “either agrees with reality or it doesn't.” -
BLOCKER— 3. Correct the category history and the “earliest milestone” claim — Exact claim: “The real lineage runs through precision metrology, not woodworking”; Maudslay's circa-1805 micrometer is “the earliest solidly-dated milestone”; and no museum record of an actual antique straightedge was located. -
SHOULD-FIX— 4. Separate the flip test, three-straightedge comparison, and Whitworth three-plate method — Exact claim: “The same self-verification logic survives today” in the flip test, and three mutually agreeing straightedges are treated as Whitworth's three-plate logic “unchanged.” -
SHOULD-FIX— 5. State what a straightedge actually samples when checking flatness — Exact claim: The tool checks “flatness of a board face” and tells whether a surface “is flat,” while one later sentence correctly says a tool proven straight along its edge says nothing about an entire surface. -
SHOULD-FIX— 6. Bound the light-gap and feeler-gauge method — Exact claim: “The thickness of the largest gauge that slides in without forcing indicates the deviation at that point,” and the Lee Valley adjacency of feeler gauges is treated as confirmation of the method. -
SHOULD-FIX— 7. Correct the granite straightedge tolerance units — Exact claim: Granite straightedges are listed as laboratory grade AA “~.000025\"/6\"” and inspection grade A “~.000050\"/6\".” -
SHOULD-FIX— 8. Make DIN 874 claims part-, form-, length-, and grade-specific — Exact claim: “DIN 874 formalized machinist-grade tolerances into four graded classes,” followed by isolated values such as grade 00 “better than ~2.33 µm over 200 mm” and grade ranges expressed in inches without a complete table. -
SHOULD-FIX— 9. Repair the Jim Tolpin / Richard Maguire attribution and quotation trail — Exact claim: SE11 attributes “Let's Get This Straight” and its comment thread to Richard Maguire/The English Woodworker, while the references say it is “corroborated” by By Hand & Eye and the open-gaps section says the By Hand & Eye author could not be confirmed. -
SHOULD-FIX— 10. Downgrade shop-made wooden-straightedge theories to their actual evidentiary level — Exact claim: A wooden straightedge is “typically rift-sawn,” tapered to “expose more end grain for faster moisture equilibration,” and finished with tung oil plus a shellac vapor barrier as though these are established design requirements. -
SHOULD-FIX— 11. Correct company continuity and volatile market evidence — Exact claim: Starrett provides “unbroken 140+ year commercial continuity for a specific product line,” and July 2026 prices/statuses are presented as proof of existence. -
SHOULD-FIX— 12. Lower confidence and repair the evidence hierarchy — Exact claim: Effective confidence is MEDIUM-HIGH because of an “unusually well-documented ancestral chain,” a forum thread, current product pages, and “unbroken” company continuity.