All charts

HELD, AND THE REASON MATTERS. The design draft carried a MAX AMPS column on the wire gauge page reading 35 A at 14 AWG, 47 at 12, 60 at 10, rising to 380 at 4/0. Those figures match the "maximum amps for chassis wiring" column of the hobby AWG table that has circulated since the mid-century, which is free air rather than raceway. NEC 310.16 for copper at 75 °C gives 20, 25 and 35 A for the same three sizes, and 240.4(D) then caps protection at 15, 20 and 30 A. The draft figures ran roughly double the code limit, on the page a person sizing a receptacle circuit would land on first. A footnote saying "depends on installation method" does not undo a printed number. Transcribe 310.16 and 240.4(D) directly, keep the two temperature columns separate, and do not merge this into the wire gauge page, because the whole point is that ampacity carries its own conditions.

HELD, and it is a near miss worth naming. Tap drill for a coarse metric thread follows the rule drill = major diameter − pitch, which reproduces the draft values exactly (M6 × 1.0 → 5.0, M8 × 1.25 → 6.8, M10 × 1.5 → 8.5). That makes the tap drill column derivable, but only from the PITCH column, and the pitch series itself is a published table from ISO 261 rather than arithmetic. A chart is only as derived as its least derivable column. Transcribe and sight-check the pitch series first; the tap drill column can then be computed and machine-checked.

HELD, and worth doing early: `number drill bit size chart` measures 880/month at KD 16, one of the cheapest entries on the page map. The number series is a published table rather than arithmetic: the sizes do not follow a formula, which is exactly why it needs a sight-check against ANSI/ASME B94.11M rather than a clever generator.

HELD. Published table, no formula. Small on its own but it completes the drill hub, and a hub that names four series and delivers two is only half honest.

HELD. The preferred metric series is a published selection, not every 0.1 mm step, so the coverage note has to say which it is before this can go up.

HELD, and this is the one to build first after the sight-check capacity exists: $7.47 CPC, the highest on the page map. The trap is the one the column model already handles: three materials, three different gauge standards, so each material column must carry its own source row.

HELD. The design draft was headed GRADE 5 (ZINC) while the figures read closer to dry unplated, and zinc-plated fasteners take less torque than dry unplated for the same tension. The draft was also internally consistent, with lubricated holding at 73–77 % of dry across every row, which again proves only consistency. Do not ship a single torque column: give each thread condition its own column with its own condition line, so the heading can never disagree with the numbers. Verify separately for J429 grades 2, 5 and 8, and treat metric ISO 898-1 as a different page.

HELD. Pitch series are published tables. This one is worth doing early anyway: the cluster measured 0 % AI Overview, which is the cleanest signal on the page map.

HELD. The design draft showed -006 through -012 with IDs of 0.070, 0.103, 0.139, 0.210, 0.239, 0.301 and skipped -009 without a note. The values look like the series shifted by one row: the draft -011 (0.239) and -012 (0.301) sit where -010 and -011 belong. The draft -006 also had ID equal to cross-section, which is not how the start of the series reads. Note that this table passed every mechanical check available, converting inches to millimetres perfectly on every row, which is exactly why a sight-check against the standard is not optional. Read AS568 directly, decide whether the page carries the full dash-number run or a labelled selection, and never leave an unmarked gap in a series.

HELD. The socket list itself is conventional and easy to enumerate, but the useful column is which bolt or nut each size fits, and that comes from the width-across-flats dimensions in ASME B18.2.2. Enumerating the sockets without that column would be a page with nothing on it that a reader could not guess.