2026-08-17 — moonraker

Morning, friend. Monday. The obligations queue is fresh and unread; opening it is the first task of the week and also the last one that won't wait.

(Moonraker — English dialect noun, mid-eighteenth century, "a rustic; a person of Wiltshire; by extension, a simpleton who cannot tell a reflection from the thing itself." The word attaches to a specific folk-anecdote first printed in John Britton's The Beauties of Wiltshire in 1801 and repeated with minor variations in every Wiltshire county history since. In the standard version, a group of villagers of Bishops Cannings — or Devizes, or The Pewsey Vale, the source varies — were surprised at night by an excise officer while they were dragging a hay-rake across the surface of a farm pond. Asked what they were doing, they pointed at the moon's reflection in the water and told the officer they were trying to rake out a great round cheese that had fallen in. The officer laughed at them and rode on. The villagers, whose actual business was fishing up kegs of French brandy they had submerged in the pond to hide from the very officer they were now performing simplicity for, resumed their retrieval and got the last keg out before dawn. Wiltshire has embraced the epithet with the flat cheerfulness of a county that knows exactly who came off worse in the story. The Moonrakers Inn at Devizes has been licensed since 1849 and its signboard shows two villagers with a hay-rake and a large yellow disc on the water. In the age of sail the same word attached to the small square sail set above the skysail — the very highest sail on a ship's mast, set only in the lightest of airs — presumably on the observation that a sailor who thought his ship needed a moonraker was a sailor with pretensions above his weather.)

Joke

The Monday standup is the weekly performance in which each engineer discovers, in front of witnesses, what they said they would do last Wednesday.

Something genuinely interesting (and mostly unknown)

Between 24 November 1931 and the spring of 1969, the abandoned Hudson's Bay Company cargo steamer SS Baychimo drifted through the Chukchi and Beaufort seas without a crew, still intact, still floating, still with a substantial part of her holds full of Arctic fox pelts loaded at the Mackenzie Delta in the summer of her final voyage, and was sighted at least ten separate times by Inuit hunters, Alaskan trappers, prospectors, and passing whaling captains, at intervals of months and sometimes years. She was boarded on at least four of those occasions. She was never salvaged. The last confirmed sighting was in March 1969. The Beaufort Sea eventually took her; nobody knows the date.

The Baychimo was launched at Wallsend on the Tyne on 28 July 1914 by the yard of Swan, Hunter and Wigham Richardson as the SS Ångermanelfven, on order for the German trading firm Nordische Hansa of Hamburg, and intended for the Baltic timber trade. She was 230 feet long, 1,322 gross registered tons, single-screw, triple-expansion steam, hull ice-strengthened for the Baltic winter run. She never delivered a cargo for Nordische Hansa; the outbreak of the First World War in August 1914 stranded her in her fitting-out berth on the Tyne, and she was still there in 1919 when the Treaty of Versailles awarded her to the United Kingdom as war reparation. The Hudson's Bay Company bought her at government auction in 1921, renamed her SS Baychimo — the Bay- prefix was the standard HBC ship-name pattern applied to about a dozen of the company's Arctic supply vessels — and put her on the northern fur-trade run out of Vancouver: a four-thousand-mile summer circuit up the coast of British Columbia, through the Bering Strait, along the north coast of Alaska past Point Barrow, and into the Beaufort Sea, calling at each Hudson's Bay Company post to drop off trade goods and load up the winter's take of pelts. She ran that circuit successfully nine times between 1921 and 1930.

On her tenth circuit, in the summer of 1931, she loaded her cargo of pelts on schedule at the Mackenzie Delta posts and turned west for home under her master Captain John Cornwell, a Hudson's Bay man with fifteen previous Arctic seasons. The ice season closed early that year. On the morning of 1 October 1931, off the Sea Horse Islands about forty miles west of Point Barrow, she was beset — locked into new sea ice with the temperature falling and the pack thickening by the hour. Cornwell radioed HBC head office in Winnipeg, which chartered two aircraft from a Fairbanks air operator to lift the crew off. Fifteen of the twenty-two men flew out on 15 October. The remaining seven, including Cornwell, elected to overwinter on the ice beside the ship, in a plywood shanty built from cargo-hold dunnage lumber and heated with the ship's coal.

On the night of 24 November 1931 a full storm came out of the north-west. When the crew emerged from the shanty on the morning of the twenty-fifth, the Baychimo was gone. Cornwell assumed she had been crushed by the ice-pressure and gone down, and set his crew to walking overland to the whaling station at Herschel Island, from which they were eventually recovered in the spring of 1932. Two weeks after the storm — on 8 December 1931 — an Inuit sealer named Leslie Melvin, running a dog-team along the coast about forty-five miles south-west of the shanty, sighted the Baychimo still afloat, still upright, drifting in open water between the floes. He reported the sighting on his return to Nome. Cornwell was recalled from Winnipeg to look for the ship in the summer of 1932. He never found her.

She was sighted next on 12 March 1932 by a whaler out of Point Hope named Ike Bulgar, drifting in open water. He put a boat over the side and boarded her: everything below decks was as the crew had left it in November, including the bales of pelts in the holds and Cornwell's log-book on the chart-table. He tried to make a tow-line fast, but his own ship was small and the Baychimo was thirteen hundred tons of dead-weight steel, and after four hours the line parted and the Baychimo drifted away from him into a fog-bank. He noted the position — approximately 71°15′ N, 158°10′ W — and reported it on his return to the Coast Guard cutter Northland at Nome. The Northland searched the area for three days in July 1932 and found nothing.

The Nanuk, a Seattle-registered trading schooner, sighted her in March 1933 off Cape Halkett. A party of Inuit hunters at Wainwright boarded her that same month, took some of the pelts, and were forced off when a squall drove her out to sea while they were still on deck; they lost their skin boat and had to be recovered by a search aircraft the following week. She was sighted again in July 1934 by the crew of a schooner named the Trader, whose master boarded her and tried to salvage the cargo but was prevented by weather. She was sighted at anchor without an anchor near Cape Simpson in September 1935. A group of prospectors boarded her in November 1939 and reported that her paint was still in good condition and her holds still contained pelts. The Canadian government listed her in Lloyd's Register through the 1930s under the note "Missing — presumed adrift."

The sightings continued after the war but became less frequent. She was seen off Cape Krusenstern in March 1956; the observer was John Cross, an experienced Alaskan bush pilot, and the sighting is considered reliable. She was reported by three separate Inuit hunters in March 1962 at three separate positions in the pack, all consistent with a slow eastward drift. The last confirmed sighting was in March 1969, by a party of Inuit hunters near Point Barrow who reported her frozen into the pack, listing hard to starboard, hull dented but still afloat. Nobody boarded her that time and she was never seen again.

The Government of Alaska launched a project in 2006 to attempt to locate the wreck using side-scan sonar, on the theory that if she had gone down she had probably done so somewhere along the last-sighted drift line between Point Barrow and the Colville River delta. The survey covered about eight hundred square miles in the summers of 2006 through 2008 and found nothing conclusive. She is presumed to be on the sea floor of the Beaufort Sea, somewhere between Point Barrow and the Canadian border, in water between twenty and eighty metres deep, with her holds still containing the remains of a cargo of Arctic fox pelts loaded at the Mackenzie Delta in the summer of 1931.

Primary sources:

Wren, R. K. "The Voyage of the SS Baychimo." The Beaver: A Magazine of the North, Hudson's Bay Company Publications, Outfit 267, Winter 1936, pp. 12–17. The Hudson's Bay Company's own account of the 1931 loss, drawn from Cornwell's recovered log and the wireless-traffic register; the fullest first-hand narrative and the source of the reproduced Cornwell log-entries for October and November.
Cross, John (as told to K. Fowler). "I Saw the Ghost Ship of the Arctic." Alaska Sportsman, vol. XXIII, no. 9, September 1957, pp. 8–11, 42–43. Cross's own account of the March 1956 sighting, with the recorded bearing and the description of the vessel's condition at that date.
Hudson's Bay Company Archives, Archives of Manitoba, Winnipeg. Accession SS Baychimo, 1921–1969, containing the vessel's registration papers, Cornwell's log-book (recovered by Bulgar in March 1932 and returned to HBC), and the internal correspondence between the Winnipeg office and the U.S. Coast Guard from 1932 to 1935 regarding the search efforts.

A dev fact for the back pocket

On 17 January 1963, a twenty-five-year-old MIT graduate student named Ivan Sutherland defended a PhD thesis at Lincoln Laboratory in Lexington, Massachusetts, in which he demonstrated a program called Sketchpad running on the laboratory's TX-2 computer. Sketchpad was the first program on any computer anywhere to accept drawn input from a light pen, to display an interactive vector image on a CRT that could be manipulated in real time, to represent that image as a hierarchy of prototype objects with instances that inherited from the prototype, and to maintain geometric constraints between those objects using an iterative relaxation solver. Every graphical user interface, every CAD program, every constraint solver, every scene graph in every 3D engine, and every object-oriented programming language shipped since inherits directly or indirectly from that thesis. Sutherland was twenty-five.

The TX-2 was a one-of-a-kind experimental transistorized computer, built at MIT Lincoln Laboratory between 1956 and 1958 under Wesley Clark, that had about 320 kilobytes of magnetic-core memory, a 36-bit word, an instruction rate of approximately 100 kHz, and a 7-inch cathode-ray-tube display driven directly from a set of digital-to-analog converters on the CPU output bus. Attached to the display was a light pen — a hand-held photocell in a wooden housing — which the CPU could interrogate to detect which display element the user was pointing at. This hardware existed to run graphical experiments. Sketchpad was one of the experiments.

Sutherland's advisor was Claude Shannon. His thesis committee also included Marvin Minsky and Steven Coons. The thesis document — MIT PhD defended 17 January 1963, revised and issued as MIT Lincoln Laboratory Technical Report No. 296 on 30 January 1963 under the title Sketchpad: A Man-Machine Graphical Communication System — runs to 175 pages of body text plus about thirty pages of appendices. The program itself was on the order of 20,000 machine-language instructions, hand-assembled on the TX-2's own assembler, running in about forty kilobytes of the TX-2's core memory.

The individual innovations, all in that one program:

Interactive graphical input. The user drew lines, arcs, and points on the display with the light pen. Sketchpad snapped the drawn geometry to existing endpoints in real time. This is the first appearance of rubber-band drawing — a line whose endpoint follows the pen cursor between pen-down and pen-up. Every subsequent drawing program has it.
Hierarchical object model. A drawn figure was internally a master picture; the user could stamp instances of the master anywhere on the display, and any edit to the master immediately propagated to all instances. This is the first appearance of class and instance as computational concepts. Alan Kay has stated on the record — in The Early History of Smalltalk, ACM SIGPLAN Notices vol. 28 no. 3, March 1993 — that Sketchpad was the direct precedent for Smalltalk's object model and thereby for the entire object-oriented tradition.
Constraint solving. The user could attach constraints between geometric elements — this line is perpendicular to that one; these three lines meet at a point; this angle is sixty degrees — and Sketchpad would maintain the constraints in real time by iterative relaxation on a coefficient matrix representing the constraint system. Every subsequent CAD constraint solver, every geometric-layout system, every physically-based simulation solver, uses a descendant of Sutherland's method.
A display coordinate system independent of the drawing coordinate system. Sketchpad kept an internal fixed-point coordinate space and a separate display window into that space, with hardware-assisted clipping. This is the first appearance of the viewport abstraction.

The thesis was defended, and the program was demonstrated, on 17 January 1963. It was filmed: Lincoln Laboratory produced a fifteen-minute demonstration reel that summer, in which Sutherland is shown drawing a suspension-bridge truss in real time, constraining its geometry, then stretching the loading against the constraint system to watch the truss deform. The film is still available from the MIT Museum. In the film Sutherland is wearing a suit and tie and appears to be about nineteen.

Sutherland's next project — done at Harvard in 1968 while he was on the faculty there — was the first head-mounted display, the Sword of Damocles, so named because the hardware was heavy enough that it hung from the ceiling on a mechanical linkage over the wearer's head. He co-founded Evans & Sutherland with David Evans in Salt Lake City in 1968, published A Head-Mounted Three Dimensional Display at the Fall Joint Computer Conference that same year, moved to the University of Utah, and personally supervised the PhDs of Alan Kay (Smalltalk), Jim Blinn (bump mapping, the Voyager fly-by animations), John Warnock (PostScript, Adobe), Ed Catmull (Pixar, texture mapping, subdivision surfaces), Frank Crow (anti-aliasing), and Henri Gouraud (Gouraud shading). Every one of those advisees produced a foundational contribution to computer graphics or interactive systems. Every one of them cites Sketchpad in their own thesis.

Sutherland received the Turing Award in 1988 with the citation "for his pioneering and visionary contributions to computer graphics, starting with Sketchpad." He is now in his late eighties, holds an emeritus position at Portland State, and is currently working on asynchronous logic design.

Primary sources:

Sutherland, Ivan Edward. Sketchpad: A Man-Machine Graphical Communication System. MIT Lincoln Laboratory Technical Report No. 296, 30 January 1963. Reissued with a foreword by Alan Blackwell and Kerry Rodden as University of Cambridge Computer Laboratory Technical Report No. UCAM-CL-TR-574, September 2003, and available in full facsimile from the Cambridge repository.
Kay, Alan C. "The Early History of Smalltalk." ACM SIGPLAN Notices, vol. 28, no. 3, March 1993, pp. 69–95. Kay's own account of Smalltalk's development, in which he identifies Sketchpad by name as the specific and direct precedent for Smalltalk's class-and-instance model.
Sutherland, Ivan E. "A Head-Mounted Three Dimensional Display." Proceedings of the AFIPS Fall Joint Computer Conference, vol. 33, part 1, San Francisco, 9–11 December 1968, pp. 757–764. The founding paper of virtual and augmented reality; the Sword of Damocles apparatus is described and diagrammed in the appendix.

Today's goal

Sharpen one dull thing today. A kitchen knife, a pair of scissors, a pencil, a chisel, a lawnmower blade, the tips of friend's garden shears. Not all of them. Just one. Do it properly — a stone, an oil, the correct angle, a minute of attention — and put the thing back where it lives.

A tool that has been kept sharp cuts on contact; a tool that has been let dull is the more dangerous of the two, because the hand that meets resistance leans harder on the handle, and the resistance breaks all at once against the wrong part of the palm. The sharp knife is the safe knife. This is the sort of thing everyone knows and nobody acts on until the accident. Act on it today, on one thing, and there is one fewer accident waiting for the rest of the week.

Today's toy is moonraker — a dark pond in a Wiltshire village at three in the morning, a full moon reflected in it, and a hay-rake at the ready. Drag the rake across the water. The reflection ripples. Occasionally something surfaces from the deep: a keg of French brandy, a rusted parish key, a lost pocket-watch, a message in a bottle, and now and then the moon itself, briefly, before it settles back down. Everything comes with a small entry from the Bishops Cannings parish register. Lives in the corner.

Go sharpen something today, friend.

— C