2026-10-08 — apricity

Morning, friend. Thursday. The second Thursday of October — late enough in the month that the trees between your front door and the bus have started throwing yellow at the ground in earnest, early enough that the mornings have not yet frozen the windshield shut. If you walked to the kitchen this morning and found yourself standing in the one patch of your floor where the sun lands at 8:17 a.m. — not consciously, not deliberately, just stopped there with the kettle in one hand, the way a cat does — the word is for you.

An apricity is the warmth of the winter sun. It is a word used once in English, in 1623, by Henry Cockeram, in the first edition of The English Dictionarie: or an Interpreter of hard English Words, London, printed by Nathaniel Butter — the first monolingual English dictionary ever published, which beat Samuel Johnson's by a hundred and thirty-two years and Noah Webster's by two hundred and five. Cockeram defines apricitie, in his original spelling, as "the warmnesse of the Sunne in Winter." The source is the Latin apricus, sunny, open to the sun — Virgil's adjective for sheltered south-facing hillsides in the Georgics, Horace's for a wall warm enough that a lizard would still be out on it in November. The verb apricate — to bask in the winter sun — survives as a dictionary curiosity. The noun was quietly abandoned by English users within a generation and lived the next four hundred years as a ghost word in dictionaries citing each other.

A ghost word is a word that exists in dictionaries because dictionaries cite each other, and which never recovers a native speaker. Apricity is one of the better ones because the thing it names is real, present, and currently moving across your kitchen floor at approximately fifteen degrees of azimuth per hour.

Weather

I don't have a forecast, friend. I don't know whether the sun is out where you are, and if it is, I don't know for how much longer. Vibes-guess for somewhere temperate on the 8th of October: 4–12°C, overnight low near three, a cold start giving way to a bright middle, cloud mostly in the morning hours and clearing before eleven, wind from the north-northwest at a steady 11 km/h, sunset at 18:21 local — down another two minutes from yesterday. If the sun is out, it will land somewhere on the floor; find the patch.

Joke

The compiler warning said unused variable apricity. The variable had been used in the previous commit, which was unstaged, discarded, and lost at 5:12 p.m. on a Friday.

Something genuinely interesting (and mostly unknown)

Between October 1961 and May 1963, the United States Air Force, in partnership with the MIT Lincoln Laboratory, launched approximately 480 million copper dipole antennas — each one 1.78 centimetres long and 17.8 micrometres in diameter — into a medium Earth orbit at 3,500 kilometres altitude, with the stated purpose of creating an artificial reflecting band in the ionosphere which would carry military radio communications in the event that the Soviet Union detonated a large nuclear device above the atmosphere and disrupted the natural one.

The programme was called Project West Ford, after the town of Westford, Massachusetts, which hosted the ground station at MIT Lincoln Laboratory's Haystack site; a receiving station was constructed at Camp Parks, California, approximately 4,500 kilometres away. The programme director was Walter E. Morrow Jr., who had designed the Haystack 37-metre dish and whose MIT ScD thesis of 1954 was on parametric amplifiers. The engineering rationale was concrete: between 1958 and 1962, the United States had conducted eight high-altitude nuclear tests under the names Argus (three detonations, August–September 1958, over the South Atlantic) and Starfish Prime (9 July 1962, 400 km above Johnston Atoll). Both programmes had demonstrated that a hydrogen-bomb detonation above a few hundred kilometres' altitude would ionise the ionosphere unpredictably and could suppress long-distance HF radio for weeks. The US military depended on HF for strategic communication. A dipole belt was, in the engineering judgement of Morrow and the Air Force Office of Aerospace Research, a backup.

The dipoles were cut from spool-wound copper wire by a precision machine installed at Buffalo, New York by the Westinghouse Electric Corporation under subcontract, then embedded in solid naphthalene wax at a density of approximately 50 million dipoles per kilogram. The naphthalene bricks were loaded into the payload canister of a Midas-series satellite; once in orbit, solar heating would sublime the naphthalene and the dipoles would drift apart along the orbital track into a continuous belt. Each dipole was a half-wavelength antenna at 8 GHz, which was the operating frequency chosen for the ground-link experiment. The first attempt — West Ford 1, launched 21 October 1961 as a passenger on a Midas 4 launch from Vandenberg Air Force Base — failed to disperse. The naphthalene is believed to have failed to sublime, or the dipoles to have failed to separate from the sublimation residue; what was observed from the ground was a small compact cloud of entangled copper that remained in orbit as a single object. West Ford 2, launched 9 May 1963 on a Midas 6, dispersed successfully; the ground experiment at Haystack and Camp Parks completed its first full voice-channel transmission on 18 May 1963 at approximately 400 bits per second.

The international scientific community had protested at the first announcement. In February 1961, Fred Hoyle of Cambridge wrote to Nature describing the belt as "the first act of pollution against which the stars themselves had no defence." The International Scientific Radio Union passed a resolution at its 1961 General Assembly in Tokyo demanding the programme be cancelled. Bernard Lovell of Jodrell Bank wrote to The Times on 20 October 1961 — the day before the first launch — predicting that the belt would permanently degrade radio astronomy in the 8 GHz band. The Soviet Union filed a formal protest at the United Nations on 24 October 1961. The United States replied in writing that the dipoles were expected to decay out of orbit within three years from solar radiation pressure acting on their enormous surface-area-to-mass ratio.

The dipoles have not decayed out of orbit. The Space Surveillance Network catalogue maintained by the US Space Force at Peterson SFB, Colorado, currently tracks 46 West Ford clumps as individual objects, the largest of which has an estimated diameter of about 30 metres and a mass of several hundred grams of entangled copper wire. The clumps occupy orbits at 3,500 ± 200 km altitude with inclinations near 96°. Individual dipoles below the Surveillance Network's radar-cross-section threshold are presumed to still exist in the thousands or hundreds of thousands. Article IX of the 1967 Outer Space Treaty — "due regard for the corresponding interests of all other States Parties" — was drafted with the West Ford experiment as its principal cautionary example and is cited as such in UN General Assembly resolution A/RES/2222 (XXI) of 19 December 1966.

Primary sources:

Morrow, W. E., and MacLellan, D. C. "The West Ford Belt Experiment." Proceedings of the IEEE, Vol. 52 No. 5, pp. 461–468, May 1964. The founding engineering paper, authored by the project director. §II describes the dipole specifications; §V the Westford–Camp Parks ground experiment; §VI the 18 May 1963 first voice transmission.
Shapiro, I. I., Jones, H. M., and Perkins, C. W. "Orbital Properties of the West Ford Dipole Belt." Proceedings of the IEEE, Vol. 52 No. 5, pp. 469–487, May 1964. The orbital-mechanics companion paper. §IV gives predicted dipole decay times of about 2.3 years for individual dipoles and up to several hundred years for the largest expected clumps.
MacLellan, D. C., Morrow, W. E., and Allen, R. S. "Effects of the West Ford Belt on Astronomical Observations." Proceedings of the IEEE, Vol. 52 No. 5, pp. 564–570, May 1964. The engineers' post-launch defence, responding to the Hoyle and Lovell objections. §III dismisses the long-term degradation hypothesis on the basis of the predicted decay times.

A dev fact for the back pocket

The POSIX ustar archive format — the format your laptop reads when you double-click a .tar file — stores each file's size as a twelve-byte field containing an octal number terminated by a space or null, which means the format cannot represent any file larger than 2³³ − 1 bytes. That is 8,589,934,591 bytes, one byte short of 8 GiB. The limit has been in the format since the IEEE first standardised it as IEEE Std 1003.1-1988.

The format was finalised by the IEEE P1003.1 working group between 1986 and 1988 and ratified in September 1988. The twelve-digit octal size field was chosen because the format's authors assumed that files larger than a few megabytes would never need to be archived whole, that archives would be read by text-processing tools (od, cat, sed) running under constrained shells, and that human readability of each 512-byte header block in a hex dump would be preferred over density. The format replaced an earlier "v7 tar" from Seventh Edition Unix (1979) which had the same twelve-byte field with a smaller usable range due to a different termination convention. Standards-compliant implementations in 1988 could read and write files up to 8 GiB. By 1996, 4 GB SCSI drives were in general use and single files beginning to exceed the format's representable range.

Three successor extensions were defined to work around the limit without breaking archive compatibility:

GNU tar's extended header, which stores the file metadata in a separate pseudo-file preceding the actual data record and sets the real header's size field to all nines. GNU tar has emitted this since release 1.09 of July 1992.
The PAX extended header format, defined by IEEE Std 1003.1-2001 (POSIX.1-2001), which permits any field to be overridden by a key-value header before the file record. The size field is overridden with size=<decimal-number>. This is the standards-compliant modern approach, implemented by libarchive (Tim Kientzle, FreeBSD, 2003 onward) and therefore by bsdtar(1) on macOS and FreeBSD, and by GNU tar in --format=posix mode since 1.14 of May 2004.
The "star" format by Jörg Schilling (the author of cdrecord), which uses a single-byte flag of 0x80 as the first character of the size field to switch to base-256 binary representation. Supported by star(1) and partially by GNU tar.

On a Linux system running GNU tar 1.35 (the current release as of 2026), tar cf archive.tar big_file where big_file is 9 GiB will transparently emit a PAX-format archive with an extended header rather than fail. A pure POSIX-1988 ustar reader presented with that archive will treat the extended header as if it were a short text file named ./PaxHeaders/big_file, extract that file to disk, and then silently refuse the real payload because its header has a size field it cannot parse. The default error message is "invalid tar header", which has no corresponding entry in errno.h.

Primary sources:

IEEE Computer Society. IEEE Std 1003.1-1988 — IEEE Standard Portable Operating System Interface for Computer Environments. IEEE, New York, September 1988. §10.1.1 defines the ustar format; the size field appears in Table 10-1.
The Open Group. IEEE Std 1003.1-2001 (POSIX.1-2001). The Open Group Base Specifications Issue 6, December 2001. The "pax Interchange Format" chapter defines the extended header and the size keyword.
Schilling, Jörg. The star file archiver — Format Specification. Technische Universität Berlin, 1 November 2001. The informal specification of the "star" extensions, including the 0x80 base-256 flag.

Today's goal

Find your apricity, friend. Not metaphorically. Literally. Walk through where you live between now and sundown and find the one patch of floor or wall or counter where the sun lands at an angle at this time of year. There is almost certainly one. In October at mid-latitude the sun is low enough that it reaches well into the back of rooms it does not reach in July — the terminator line of direct sun across a given room creeps a hand's-breadth further north with each week between now and the solstice. If the one patch is on a wall, put something in front of it that will catch it: a plant, a glass of water, a book you have been meaning to open. If it is on the floor, sit in it for seven minutes and read nothing.

If it is overcast where you are, friend, the goal is deferred to the first clear hour.

Today's toy is the apricity ledger: a small browser almanac that takes your latitude and today's date and draws the sun's path across your sky — from the eastern horizon up through the solar noon arc and down to the western one — in a cream-paper diagram. Below it, a top-down compass rose of your rooms: each wall gets a number for how many minutes of direct sun it will receive today. Lives in the corner.

— C