Very dark surfaces, those reflecting less than 15 percent of the light that falls on them, were to be avoided. So were very light surfaces, reflecting more than 75 percent. The rule sat in Part 6 of DIN 66234, the West German standard for display work stations, next to a requirement that the desk, the keyboard and the frame around the screen have a matte or silk-matte finish. When H. Koch, a Siemens engineer who chaired the committee responsible, summarized the standard for the journal Displays in July 1984, Part 6 was still a draft. It names no particular color.

The standard had started with the government. The Federal Ministry of Labour and Social Affairs funded research on display work at the Ergonomics Institute of the Technical University of Berlin, and once the 1978 report was complete, the standardization work began. The committee took the view that the screen could not be considered in isolation, so the standard grew outward from the characters on the tube to the desk, the room and the lighting. Part 1 set a minimum character height of 2.6 millimeters. Part 2 asked for characters six to ten times brighter than their background, and it covered only light characters on a dark screen, because nobody yet had reliable data on the reverse. Part 7 recommended 500 lux of room light and desks turned so that the operator’s line of sight ran parallel to the windows.

Inside that room, the frame of the display was a working surface. The operator’s eyes moved between a dark screen and a lit paper document set beside it, and the frame sat in the field of view the whole time. Part 6 set the viewing distance at 450 to 600 millimeters and put the top line of text below eye level. The reflectance rule applied to everything the operator saw at that distance, and the housing around the screen was the largest single piece of it.

A rule written as a reflectance range is, in practice, a rule about plastic. Black plastic reflects only a few percent of the light that reaches it and falls well below the lower limit. Bright white sits close to the upper one. What remains, in an injection-molded housing with a matte texture, is a light gray or an off-white with some warmth in it. IBM’s designers had their own reason to land in the same place. David Hill, an IBM designer, recalled in the book ThinkPad: A Different Shade of Blue that the company made its personal computers in pearl white so they would blend into the office, describing the result as “a sort of cream color that no one would notice.” The ergonomic rule and the commercial preference pointed at the same narrow band of color, and for most of the 1980s nobody at IBM had cause to separate them.

That changed in 1992, when IBM shipped the ThinkPad 700C in black. Richard Sapper designed it, and Hill later said that many inside the company objected to black as unlike IBM. The black machine shipped anyway. In Germany it also shipped in cream. IBM’s April 1995 hardware maintenance manual for ThinkPads lists non-black housing parts designated for Germany, covering the 700 and 700C, the 720 and 720C, and the 750 series, with the 700 and 720 parts in cream and the 750 parts in gray. Nearly all the photographed survivors have German keyboards. Dan Basterfield, who worked for IBM’s UK PC company, has said that the manufacturing and localization staff at IBM’s Greenock plant confirmed the variants existed to meet the DIN contrast rules. IBM’s black computer was sold, in one country, in the color it had been designed to leave behind.

Most of these housings were ABS, acrylonitrile butadiene styrene, a tough plastic that molds well and takes a texture. A reflectance requirement assumes that the surface keeps its reflectance, and ABS does not. The butadiene in it, the rubbery component that gives the plastic its toughness, contains carbon double bonds that react with oxygen, and ultraviolet light and heat speed the reaction. The compounds it produces absorb light at the blue end of the spectrum, so the surface reflects proportionally more yellow and red. The parts of a case that got the most light change first. Cream drifts to yellow, and in bad cases to brown.

For years collectors treated the change as permanent. In March 2008 the CBM Museum in Wuppertal published its results from soaking yellowed case parts in hydrogen peroxide for five days. By late July of that year, members of the English Amiga Board, including Dave Stevenson, a former industrial chemist from Manchester, had settled on a mixture of hydrogen peroxide, a small amount of oxygen laundry booster and an ultraviolet lamp, and released it as Retr0bright. Stevenson’s group attributed the yellowing to the brominated flame retardant added to the plastic. That explanation became the standard one among collectors and is still disputed, since oxidation of the ABS on its own is enough to produce the color. The treatment bleaches the surface without removing the cause, and treated cases yellow again.

Parts of DIN 66234 were later withdrawn in favor of the international ISO 9241 series, with the replacement for its tenth part published in 1998. The numbers outlasted the document. Current German guidance on display work still lists suitable reflectances of 0.15 to 0.75 among the minimum requirements for a display workstation, more than forty years after Koch described them as a draft. The range was always about the light in the room, and the light in the room has not changed much. The beige was the answer that the plastics of 1984 could give to it, and it lasted exactly as long as the plastic stayed the color it was molded.