Directory / Source discrepancies

Source register · 19 findings across 8 manufacturers

Where the sources disagree.

Verifying a record here means reading a manufacturer's own documents against each other. Sometimes they do not agree — a weight that is lighter with a magazine than without one, a spec sheet naming two different model families, a page whose specification panel contradicts a SKU listed on the same page. This register collects 19 such findings across 8 manufacturers and 19 records, each with what was published, why it cannot be read as-is, and the document to open to see it. Nothing here is repaired: where a maker publishes two figures that cannot both be true, both stand on the record, and where no reading survives, the field is left blank.

What this page is about, and what it is not: every entry below is a defect in a document — a published figure, a label, or a date. None of it is a claim about any firearm’s design, safety, quality, or fitness for any use, and none of it should be read as one. A manufacturer printing a magazine weight in the wrong column has made a typesetting error, not a pistol.

Figures that cannot be right06

A published value that fails on its own terms — arithmetic that does not work, or a measurement that cannot describe the thing it is attached to. These need no second document to see.

01Heckler & Koch

A pistol that weighs less with a magazine in it

Published: HK's USP Compact specification table carries two weight columns, “Without magazine” and “With empty magazine.” For the USP9 Compact they read 770 g / 27.20 oz and 727 g / 1.60 lb.

The problem: A pistol cannot weigh less with a magazine fitted than without one, so at least one cell is wrong — and the table does not say which. The usual tell is absent because both figures are internally consistent in their own units: 770 g is 27.16 oz, and 1.60 lb is 726 g. The 770 g is also byte-identical to the full-size USP9's without-magazine weight, which would be surprising for a pistol 22 mm shorter with a 17 mm shorter barrel. Every row of the table has the same defect, not just this one.

What this directory did: No weight at all is recorded on any of the three USP Compact records. The archived operator's manual cannot arbitrate — its specification table publishes no weight, no length and no barrel length. With two published figures that cannot both be true and nothing to break the tie, the honest output is a blank.

Where to check it: The USP Compact product pages on hk-usa.com, read as Web Archive captures because the live site serves a bot challenge. Each record cites its own capture.

3 records affectedFound while verifying, 2026-08-13

Heckler & Koch USP9 Compact (V1)Heckler & Koch USP40 Compact (V1)Heckler & Koch USP45 Compact (V1)

02Heckler & Koch

A 9 mm magazine listed at the weight of the pistol

Published: The full-size USP page's weight table has “Without magazine” and “Magazine” columns. The .45 row gives a sensible 98.9 g magazine. The 9 mm row gives 770 g — the pistol's own without-magazine weight, repeated — and the .40 row gives 830 g.

The problem: A 9 mm pistol magazine does not weigh 770 g, and the figure is the neighbouring cell's value. Read alongside the USP Compact pages, the defect is systematic across HK's weight columns rather than one slip.

What this directory did: It touches no recorded value: these records take only the “Without magazine” figure, and there is no schema field for magazine weight in any case. It is registered because it bears on how far the rest of the table should be trusted — and the rest was then checked. Every dimension pair on the page reconciles.

Where to check it: The full-size USP product pages on hk-usa.com, as Web Archive captures.

3 records affectedFound while verifying, 2026-08-13

Heckler & Koch USP9 (V1)Heckler & Koch USP40 (V1)Heckler & Koch USP45 (V1)

03Heckler & Koch

One weight, two units, an ounce apart

Published: Both the VP9 and VP9-B pages give 753 g and 25.56 oz under a row headed “Weight (with empty magazine).”

The problem: 753 g is 26.56 oz. The two figures HK prints for one pistol are a full ounce apart, and the digits differ by one, which reads like a transcription slip in the imperial value — but that is a guess about which figure is wrong, not evidence.

What this directory did: Both stand as published on both records, and neither is flagged as converted, because this directory converted neither: marking one would pin HK's arithmetic on the directory, and picking the correct one would be a repair rather than a record. Every other unit pair on both pages reconciles, so the defect is confined to one row.

Where to check it: The VP9 and VP9-B product pages on hk-usa.com, as Web Archive captures dated February 2025 and December 2024.

2 records affectedFound while verifying, 2026-08-13

Heckler & Koch VP9 Heckler & Koch VP9-B

04Glock

A weight row containing a length

Published: The G23's product page prints its weight-without-magazine row as “102 mm / 4.02 inch.”

The problem: That is a length, and specifically it is the page's own barrel-length value copied into the weight row. There is no reading under which a weight is expressed in millimetres.

What this directory did: Weight without magazine is left unrecorded on the G23. Glock's owner's manual publishes no without-magazine weight for any model, so nothing else on the record can supply it.

Where to check it: The G23 product page on us.glock.com.

1 record affectedFound while verifying, 2026-08-11

Glock G23 Gen 5 MOS

05Smith & Wesson

Two different magazines converted to the same wrong figure

Published: The Shield Plus specification sheet lists magazine weights of 2.6 oz for the 13-round and 2.2 oz for the 10-round, giving 6.6 g as the metric figure for both.

The problem: 2.6 oz is about 74 g and 2.2 oz about 62 g. Neither is 6.6 g, and two different imperial values cannot share one metric conversion.

What this directory did: Left unrecorded, as the G23's misprint is. Because it raises a fair question about the sheet's other metric figures, every one of them was then checked and they are all correct — 7.9 cm for 3.1 in, 15.5 cm for 6.1 in, 11.7 cm for 4.6 in, 2.8 cm for 1.1 in, 2.4 cm for 0.94 in, 13.5 cm for 5.3 in and 507.5 g for 17.9 oz. The error is confined to the two magazine rows.

Where to check it: The M&P 9 Shield Plus specification sheet for SKU 13248, linked from the Smith & Wesson product page.

1 record affectedFound while verifying, 2026-08-13

Smith & Wesson M&P 9 Shield Plus

06CZ

A weight whose two units are twelve grams apart

Published: CZ publishes the Shadow 2's weight as 1330 g and 46.5 oz.

The problem: 1330 g is 46.9 oz, and 46.5 oz is 1318 g. The gap is CZ's own, sitting between two figures printed side by side for one pistol.

What this directory did: Both are recorded as published and neither is flagged as converted, because neither was converted here. CZ is otherwise the richest structured source in this directory — every measurement carries both units — which is exactly why the one row that does not reconcile is worth naming.

Where to check it: The CZ Shadow 2 product page on czfirearms.com, SKU 91257.

1 record affectedFound while verifying, 2026-08-12

CZ Shadow 2

One maker, two documents, two answers09

A manufacturer's own sources disagree about one pistol. Where this directory has to choose, it says which document it followed and why, and cites the other so the disagreement stays checkable.

01Glock

The manual and the product page differ on two dimensions

Published: Glock's owner's manual (p. 57) gives the G29 Gen5 an overall length of 174 mm / 6.85 in and a height of 113 mm / 4.45 in. The product page gives 176 mm / 6.93 in and 115 mm / 4.53 in.

The problem: Two manufacturer documents, one pistol, 2 mm apart on both axes. Neither document acknowledges the other.

What this directory did: The record follows the current product page, the same rule applied to the G17L and G41. The manual agrees with the page on barrel length, capacity, trigger pull and weight with magazine, so the disagreement is confined to those two rows — and it was found while verifying a different record from the same manual page, which is why the checking habit matters.

Where to check it: Glock's owner's manual, p. 57, against the G29 Gen5 product page on us.glock.com.

1 record affectedFound while verifying, 2026-08-12

Glock G29 Gen 5

02Glock

Subcompact on the page, compact in the launch release

Published: The G29 Gen5 product page's structured size-class label and its overview both read “Subcompact.” Glock's January 2024 launch release calls the same pistol “a compact pistol.”

The problem: The two are different size classes in Glock's own vocabulary, and this directory normalizes its frame classes from the maker's published label — so which document is read decides which class the record carries.

What this directory did: The record follows the structured size-class label rather than the release's prose, because a specification field is a classification and a sentence in a press release is a description. The same call keeps a marketing sentence from overriding a specification row elsewhere in the directory.

Where to check it: The G29 Gen5 product page on us.glock.com, against Glock's January 2024 launch release.

1 record affectedFound while verifying, 2026-08-12

Glock G29 Gen 5

03Glock

Two millimetres between the manual and the page

Published: Glock's owner's manual (p. 59) gives the G17L Gen5 an overall length of 242 mm / 9.53 in. The MOS product page gives 240 mm / 9.45 in.

The problem: One measurement, two manufacturer documents, no acknowledgement in either that they differ.

What this directory did: The record follows the current product page. The manual agrees on width, height, capacity, trigger pull and weight with magazine, so only the one row is in dispute.

Where to check it: Glock's owner's manual, p. 59, against the G17L Gen5 MOS product page on us.glock.com.

1 record affectedFound while verifying, 2026-08-12

Glock G17L Gen 5 MOS

04Glock

The launch release and the current page give three different dimensions

Published: Glock's G41 launch release lists a 226 mm / 8.90 in overall length, a 32.5 mm / 1.28 in width and a 680 g / 24 oz weight without magazine. The current product page gives 223 mm / 8.78 in, 34 mm / 1.34 in and 675 g / 23.81 oz.

The problem: All three axes disagree, and the width by 1.5 mm — larger than the rounding either document publishes at. A launch release describes the pistol at announcement and a product page describes it now, which may explain the gap without resolving which figure a reader should quote.

What this directory did: The record follows the current page. The release is still cited on the record, because it is the only source that dates the model.

Where to check it: Glock's G41 launch release, preserved as a Web Archive capture, against the G41 Gen4 product page on us.glock.com.

1 record affectedFound while verifying, 2026-08-12

Glock G41 Gen 4

05CZ

Four millimetres of barrel between a live page and a retired one

Published: CZ's retired cz-usa.com listing gives the Shadow 2 a 4.89 in barrel and an 8.53 in overall length. The current czfirearms.com page gives 120 mm / 4.7 in and 217 mm / 8.5 in.

The problem: The overall lengths agree — 8.53 in is 216.7 mm — which makes the barrel figures the outlier: they differ by about 4 mm. And 4.89 in is the number that circulates most widely for this pistol, so the figure most often quoted is the one from the retired page.

What this directory did: The record follows the current product page and cites the archived listing so the disagreement is checkable rather than merely asserted. Nothing is averaged.

Where to check it: The Web Archive capture of cz-usa.com's Shadow 2 listing dated January 2017, against the current czfirearms.com page.

1 record affectedFound while verifying, 2026-08-12

CZ Shadow 2

06CZ

One page, three trigger-pull figures

Published: The P-10 C's technical specification table gives 25 N / 5.6 lb. The same page's model-features block renders that identical 25 N as 5.5 lb. The retired cz-usa.com listing's prose says the trigger “breaks at a clean 4-4.5 lbs.”

The problem: The first two are the same metric figure rounded two different ways on one page — 25 N is 5.62 lb, so 5.6 is the better rounding. The third is about a pound and a half away, from a different era and a different SKU. A reader quoting “the CZ P-10 C trigger pull” can find three answers on CZ's own material.

What this directory did: The record follows the technical specification table, because that is the document whose purpose is specification, and discloses the other two rather than quietly picking a favourite.

Where to check it: The CZ P-10 C product page on czfirearms.com, SKU 91531, and the archived cz-usa.com listing.

1 record affectedFound while verifying, 2026-08-12

CZ P-10 C

07Beretta

Recoil-operated in the manual, delayed blowback in the listing

Published: Beretta's 92 Series user manual states “OPERATION: Semi-automatic, short recoil of barrel” with “LOCKING SYSTEM: Falling block,” and its prose repeats that the series works “on the short barrel recoil principle and using a falling block locking system.” A Beretta USA product listing calls it “a short recoil, delayed blowback system.”

The problem: The listing's phrase is not a coherent single classification. A pistol is not both recoil-operated and blowback — those are alternative descriptions of how the action cycles, not a compound one.

What this directory did: The manual governs and the listing's wording is disclosed. The record's operating system is recoil-operated, footnoted to the manual.

Where to check it: Beretta's 92 Series user manual, captured December 2014, against the Beretta USA listing for stock number J92F300M, captured March 2012. Beretta's live site serves a bot challenge, so both are archived captures.

1 record affectedFound while verifying, 2026-08-12

Beretta 92FS

08FN

A 4.25 in SKU on a page whose spec panel says 4.0 in

Published: The FN 509 page's specification panel gives a 4.0 in barrel. The same page's PRODUCT DESIGNATION table lists seven SKUs, one of them 66-100241, “FN 509 NMS 4.25.”

The problem: A SKU whose own name states a 4.25 in barrel sits on a page whose specification panel states 4.0 in. The panel's figures cannot be carried onto it, and the page publishes no separate dimensions for that configuration.

What this directory did: The record covers SKU 66-100002 and takes the panel's figures. The 4.25 in SKU is treated as a separate listing whose dimensions this directory does not hold, rather than being given the panel's. It is the clearest case in the directory of a conflicting SKU announcing itself in its own name.

Where to check it: The FN 509 product page on fnamerica.com — the specification panel against the PRODUCT DESIGNATION table.

1 record affectedFound while verifying, 2026-08-13

FN America 509

09Heckler & Koch

A manual whose dimension column does not convert

Published: The USP9 Compact product page gives a width of 35 mm / 1.38 in. HK's archived operator's manual gives 1.14 in / 34 mm for the same pistol, 1.26 in / 38 mm for the full-size USP9, and states the USP45's height as “5.55 in / 1.41 mm.”

The problem: The manual's column does not convert: 1.14 in is 29 mm, and 1.26 in is 32 mm. The USP45 height is the clearest tell — 1.41 mm is not a height of anything, and 5.55 in is 141 mm, so a decimal point has moved.

What this directory did: The page's figure is recorded, because it reconciles exactly and the manual's column demonstrably does not. Internal consistency is the tiebreak where two manufacturer documents conflict and neither is obviously senior.

Where to check it: The USP product pages on hk-usa.com against HK's archived operator's manual, both as Web Archive captures.

3 records affectedFound while verifying, 2026-08-13

Heckler & Koch USP9 Compact (V1)Heckler & Koch USP9 (V1)Heckler & Koch USP45 (V1)

The same number under two different labels02

The subtlest kind, and the most likely to be repeated as fact: one figure appearing in two of a maker's documents under labels that do not mean the same thing. Nothing looks wrong until the labels are read.

01Smith & Wesson

The same 1.1 in called a frame width and a width

Published: The Shield Plus specification sheet publishes 1.1 in under the label “Frame Width,” with a sub-row immediately beneath reading “(Less Controls): .94 in.” The product page renders the same number under the plainer heading “WIDTH.”

The problem: A frame width and an overall width are different measurements: one describes the frame, the other the widest point of the pistol including whatever stands proud of it. The sub-row gives it away — a figure with a separate “less controls” variant is not simply the frame at its narrowest, and neither variant is announced as the pistol's overall width.

What this directory did: No overall width is recorded for the Shield Plus, and no width comparison is published against pistols whose makers do publish one. Both figures are recorded under their own labels. Where a maker labels one figure two ways, the document whose purpose is specification governs — here the sheet, which uses the more careful label.

Where to check it: The M&P 9 Shield Plus specification sheet for SKU 13248 against the Smith & Wesson product page.

1 record affectedFound while verifying, 2026-08-13

Smith & Wesson M&P 9 Shield Plus

02Magnum Research

A width in the manual, a slide width in the catalogue

Published: The Desert Eagle owner's manual lists “Width: 1.25” beside Height in a dimensions table. The Kahr Firearms Group catalogue labels the identical figure “Slide Width.”

The problem: The two documents disagree about what the measurement is rather than about its value. A slide width and an overall width are not interchangeable, and a reader comparing 1.25 in against another maker's overall width needs to know which one this is.

What this directory did: Recorded as an overall width on the manual's authority — the document whose purpose is specification rather than sale — with the catalogue's label carried in the basis note so the caveat travels with the figure. Had only the catalogue been available, the figure would not have been recorded at all.

Where to check it: The Magnum Research owner's manual against the Kahr Firearms Group 2026 catalogue.

1 record affectedFound while verifying, 2026-08-12

Magnum Research Desert Eagle Mark XIX

Documents that disagree with themselves02

Not a value in dispute but the document's own account of what it describes — which model, or when.

01Colt

A spec sheet that names two different model families

Published: The sheet's title reads “COLT GOVERNMENT SERIES 70 .45 ACP.” Its body reads “MODEL: 1991 Government.”

The problem: The Colt 1991 and the Series 70 are different model families, not two names for one pistol. The document does not say which it is describing.

What this directory did: Neither label is adopted. The record is keyed to the SKU, which the sheet states three times consistently — in the title, as “SKU: O1911C-SS” and as “MFG MODEL NO: O1911C-SS” — with a matching UPC. So the model name carries the SKU and no series or generation is recorded. The measurements survive the inconsistency because both disputed families are Government-pattern pistols at the same barrel and overall length; the series does not survive it.

Where to check it: Colt's spec-sheet endpoint at colt.com, captured March 2024. The endpoint now returns 404, so this record cannot be refreshed from its own source.

1 record affectedFound while verifying, 2026-08-12

Colt Government Model, SKU O1911C-SS

02Glock

A press release with three different dates

Published: Glock's G41 launch release is headed January 10, 2014. Its dateline reads January 6, 2013. Its body ties the unveiling to SHOT Show on January 14–17, 2014.

The problem: Three dates spanning thirteen months in one document, and the year introduced for the record depends on which is read.

What this directory did: Recorded as 2014, on the body's evidence — the SHOT Show dates are externally checkable where the header and dateline are not. The release is preserved only as a Web Archive capture, and the inconsistency is registered so the choice is visible rather than implied.

Where to check it: Glock's G41 launch release, as a Web Archive capture.

1 record affectedFound while verifying, 2026-08-12

Glock G41 Gen 4

What does not qualify: rounding is not a discrepancy. Glock publishes 86.5 mm and 87 mm barrels for the G43 and G43X and prints 3.41 in for both; Colt prints 1.30 kg where 46 oz is 1304 g. Those are the published precision doing its job. Nor is a maker declining to classify something — a silence is a gap, and this directory shows gaps as blank rows rather than as errors. An entry earns its place here only where a reader can open the cited document and see the conflict.

Found something we got wrong? This register runs outward, from this directory toward the documents it cites. It runs inward too: if a figure on any record is wrong, or a finding here is mistaken, submit a correction with the source. Corrections to this page are as welcome as corrections to a spec row, and for the same reason — a claim that cannot be checked is not worth publishing. See the methodology for how records are verified and how conflicting sources are handled.

Reference information only. Not legal, purchasing, training, carry, storage, or assembly advice. Manufacturer technical data are rounded and may change without notice, and every document cited here is quoted as published on the date stated.