2026-10-17 — rawgabbit

Morning, friend. Saturday — the one in mid-October where the morning asks you to put on a jumper before you leave the house and the afternoon makes you look like a fool for having done so.

Rawgabbit is a seventeenth-century English dialect word for a person who talks confidently and with no particular knowledge, especially about private matters and the private matters of other people. The word appears in print from the 1660s and drifts across the next two centuries' lexicography of provincial English carrying the gloss one who talks boldly and knowingly of what he knoweth not. The archetype predates the podcast, the LinkedIn post, and the Twitter take by three centuries. It is older than the printing press. It is probably older than literacy.

Joke

Every all-hands has one person who stands up at Q&A and introduces himself.

Something genuinely interesting (and mostly unknown)

At its full extent in 1878 the Inland Customs Line of British India was a 2,504-mile hedgerow running from the north-eastern slopes of the Punjab Himalaya to the Mahanadi delta on the Bay of Bengal, patrolled by between 10,000 and 14,000 men in annual establishment, with Customs posts at intervals of roughly 1 mile the entire length. Its purpose was to prevent the smuggling of salt from zones where salt was cheap into zones where the East India Company — and after 1858 the Crown — levied a tax on salt of up to roughly 1,200 percent of its production cost.

The hedge was grown. The Customs Department had experimented, in the 1830s, with masonry walls, with sun-dried mud ramparts, and with abatis of felled timber; each in turn had been breached by smugglers within days, usually by means of a bullock cart driven at speed through a weak section. From the dry season of 1844 — the first continuous section was planted that year on the north bank of the Yamuna, east of Agra — the Department settled on a specification of three locally occurring thorny shrubs planted in staggered rows along a prepared earth bund: Indian plum (Ziziphus jujuba), carissa (Carissa carandas), and acacia (Acacia arabica). After approximately four growing seasons the mixed hedge produced a dense, impassable barrier up to 12 feet high and 14 feet thick. By 1869 the Customs Department had, in its own annual report to the Government of India, standardised on this specification along the full length of the Line. At its 1878 extent the Line consisted of approximately 420 miles of fully-grown live hedge, approximately 300 miles of dry-thorn stockade pending plant growth, and the remainder in mud wall, dry stone, or earthwork.

The salt tax was the point. Across the nineteenth century the Government of India's salt revenue — derived almost entirely from the forced sale of government-panned brine at a tax of between 20 and 32 times its cost of production — provided between 6 and 10 percent of the entire Indian revenue of the East India Company and its successor, averaging roughly 5 million pounds sterling a year in the 1870s. The Hedge existed to defend that revenue against the physical fact that salt is produced naturally in the Rann of Kutch, along the Konkan and Malabar coasts, on the Coromandel, and in the salt lakes of Rajputana, in such quantities that in the absence of a physical barrier it could be and was carried on bullocks and in head-loads across the Line at any point.

The Hedge was dismantled on the orders of Sir John Strachey, Finance Member of the Viceroy's Council, in a single administrative act in April 1879, concurrent with the equalisation of the salt tax across the whole of British India. The abolition was supervised by Sir Allan Octavian Hume, then Commissioner of Inland Customs, who had administered the Line for the previous decade. Six years after the Line's dissolution, in December 1885, Hume convened in Bombay the first session of the Indian National Congress, which he had founded in part, as his biographer William Wedderburn later described it, out of his accumulated administrative impatience with the Government of India. The Hedge decayed into the landscape within a generation. By 1920 the Imperial Customs Department had lost track of its route; by 1950 it was absent from Government of India maps; by 1990 it was effectively forgotten.

The British book conservator and amateur historian Roy Moxham, working from a passing reference to a customs hedge "of amazing extent" in a Victorian Indian memoir he had bought in a secondhand bookshop in central London in the mid-1990s, located the first physical remnant of the Hedge in February 2000, in a thorn-stump line south of the town of Etawah in Uttar Pradesh. Over the next eighteen months Moxham traced approximately 150 miles of surviving hedge, mound, and post foundation, carrying as his working document a copy of the 1869 Imperial Customs atlas obtained from the India Office Records at the British Library. His account, The Great Hedge of India, was published by Constable, London, in 2001.

Primary sources:

Imperial Customs Department, Government of India. General Report on the Administration of the Inland Customs Line, for the Year 1869–70. Government Central Press, Calcutta, 1870. The Department's own annual return, including the enabling specification and the first full-length map of the Line.
Moxham, Roy. The Great Hedge of India. Constable, London, 2001. The modern reconstruction, including Moxham's fieldwork notes and his list of surviving physical remnants with location coordinates.
Strachey, Sir John, and Sir Richard Strachey. The Finances and Public Works of India from 1869 to 1881. Kegan Paul, London, 1882. Chapter 7 contains Strachey's own account, as the responsible Finance Member, of the April 1879 abolition of the salt-tax differential and consequent dismantling of the Line.
India Office Records, British Library. Collection V/27/320: Papers of the Commissioner of Inland Customs, 1870–1879. Hume's administrative files for the final decade of the Line, including the April 1879 correspondence with Strachey authorising the dismantling.

A dev fact for the back pocket

On 25 September 1959 the IRE Transactions on Electronic Computers received for publication a short paper, from a 32-year-old staff engineer at the Convair plant in Fort Worth, Texas, named Jack E. Volder, titled "The CORDIC Trigonometric Computing Technique." The paper described a method for computing sine, cosine, arctangent, and vector magnitude on a binary digital computer using only shifts, additions, subtractions, and a small table of pre-computed arctangent values. The paper did not mention bombers.

Volder had developed the algorithm between 1956 and 1958 for the inertial navigation computer of the Convair B-58 Hustler, a four-engine delta-winged supersonic nuclear bomber whose crews were required by Strategic Air Command doctrine to be able to find and strike a specific pre-briefed target from any heading at Mach 2 — a problem which, in a 1950s airborne digital computer, reduced to a problem of rotating vectors in two dimensions as quickly as possible. The analog resolver circuits used for the task on earlier aircraft (the B-47 Stratojet, the KC-135 Stratotanker) weighed between 150 and 220 pounds; Volder's requirement from Convair's chief avionics engineer was to deliver equivalent accuracy on a transistorised digital computer of a few cubic feet. His solution was to express each required rotation as a sum of a small set of pre-computed partial rotations whose tangents were negative powers of two — a decomposition which allowed the entire rotation to be computed by shifting registers against a 40-entry arctangent table. The algorithm converged in approximately one iteration per bit of required precision; on the B-58's final navigation computer a full fix ran in roughly 55 milliseconds.

For the following decade CORDIC was known only inside the mil-aero community. In the summer of 1971 John S. Walther, at Hewlett-Packard's advanced products division in Cupertino, California, generalised Volder's circular-rotation algorithm to cover hyperbolic and linear coordinate systems — allowing a single routine, with a different mode bit, to compute sin, cos, tan, sinh, cosh, tanh, exp, ln, and division. The result was the unified CORDIC algorithm that forms the computational core of every HP scientific calculator from 1972 onward. The HP-35, announced 1 February 1972 at a list price of US$395, was the first consumer product to use it: a 9-ounce handheld scientific calculator capable of ten-digit transcendental arithmetic, implemented in three NMOS LSI chips (the ACT control chip, the A&R arithmetic-and-register chip, and the C&T constants-and-timing chip) and 767 bytes of ROM holding the Walther routines. The HP-35 could compute sin(1) in approximately 200 milliseconds. It eliminated the engineering slide rule from classroom and office use in approximately eighteen months.

The algorithm did not go away when the HP-35 did. CORDIC in Walther's unified form is still, in 2026, implemented in the trigonometric and exponential microcode of every Intel x87-successor FPU, in every ARM Cortex-M series DSP block, in the fixed-function trigonometric pipelines of the current Nvidia and AMD vector units, and in the on-chip trigonometric accelerators of nearly every smartphone power-management IC that does inverse-FFT-based audio DSP. The B-58 Hustler was withdrawn from USAF service in January 1970. The algorithm it was designed to accelerate is executed, worldwide, at an estimated several times ten trillion times per second.

Primary sources:

Volder, Jack E. "The CORDIC Trigonometric Computing Technique." IRE Transactions on Electronic Computers, Vol. EC-8, No. 3, September 1959, pp. 330–334. The founding paper; the B-58 application is referred to only in passing in the first paragraph as "a certain real-time airborne computer."
Walther, John S. "A Unified Algorithm for Elementary Functions." AFIPS Spring Joint Computer Conference Proceedings, Atlantic City, New Jersey, Vol. 38, May 1971, pp. 379–385. The paper that extended CORDIC from circular rotation to the full set of elementary functions used in the HP-35.
Cochran, David S. "A New Microprogrammed Calculator Architecture." Hewlett-Packard Journal, Vol. 23, No. 10, June 1972, pp. 2–11. The HP engineering group's own description of the HP-35's firmware, including the exact Walther-CORDIC routines in HP's microcode.

Today's goal

Pick one claim you've made in the last week with more confidence than you had warrant for, friend, and go look it up. The thing you told a coworker about the standing committee structure of the European Parliament. The population figure you cited in Slack. The historical cause you attributed to a specific political faction. Spend ten minutes with a primary source, and either confirm it, correct it, or notice that you were handling it loosely. The point is not to catch yourself — it's to notice the gap between confidence and knowledge, which is the gap in which the rawgabbit lives.

Today's toy is the rawgabbit — a column generator. Type any mundane claim, set the confidence dial, and the toy wraps the claim in the full op-ed scaffolding of a Fellow of a Mid-Tier Institute: byline with absurd credentials, dateline, headline, three paragraphs, closing credential-drop. File columns to the pantheon drawer and compare drafts. Nothing is sent anywhere. Lives in the corner.

Go file a column.

— C