The Boing Ball is a red and white chequered ball that bounces across a grey and purple grid, spinning as it goes, with a deep boom each time it hits the floor. It was the first thing many people ever saw an Amiga do. Two programmers wrote the first version overnight in a back room at the Consumer Electronics Show (CES) in Las Vegas in January 1984, on a prototype that barely worked, for a company that was running out of money. The demo helped find the money, and the ball became so closely tied to the machine that it very nearly became its logo.
Fast facts
- First shown: Winter CES, January 1984, behind closed doors on the Amiga prototype, code-named Lorraine.
- Written by: Dale Luck and RJ Mical; a 1990 Amazing Computing listing also credits Sam Dicker.
- The ball: about 140 × 100 pixels, drawn once and never redrawn; the spin is colour cycling and the bounce is hardware scrolling.
- The sound: a sample of Bob Pariseau hitting a closed aluminium garage door with a foam baseball bat.
- Later life: demonstrated at the Amiga 1000’s launch in July 1985, left off the final logo, and used on Amiga branding and merchandise into the 1990s.
Written at the show
Amiga, Incorporated needed Winter CES to go well. In Jimmy Maher’s history of the machine, The Future Was Here, the company “was quickly running out of money and desperately needed to demonstrate something to the potential additional investors who might keep it afloat long enough to complete the Lorraine”. The Lorraine was not yet a computer in a case. It was “a cobbled collection of circuit boards and wires that failed constantly”, and the hardware technicians “brought piles of chips and other spare parts with them and learned to repair the machine on the fly during demonstrations”. It had no operating system and no keyboard; it “could be operated only by uploading a program from a remote terminal connection”. The programmers wrote their code on a separate 68000 computer, a Sage IV, and sent it across a cable.
They had had the prototype for only a few days before the show. Mical and Luck, Maher writes, “worked furiously at the show itself to improve upon the software the team had brought with them”, and “created the first version of the Boing demo in Amiga’s little rented back room on the CES show floor in an all-night hacking session fueled by ‘a six-pack of warm beer’”.
That first version was simpler than the one people remember. Luck recalled that at the January show the ball only bounced up and down, and made no sound. By the Summer CES in Chicago in June 1984 it moved from side to side and boomed. BYTE described the prototype demonstrations of early 1984 in its August 1985 Amiga feature: “Sound shifted from the right speaker to the left as a ball bounced across the screen. Everyone wondered if Amiga could really reduce all the power in the prototypes to silicon chips.” By the summer Atari and Commodore were both trying to buy the company, and Commodore bought it in August 1984.
How it works
The ball looks as if it is being drawn again in every frame. It is not. Maher took the demo apart for his book and found that it “operates in a fairly basic way”, without the copper lists, hardware sprites or dual playfields that later Amiga programs relied on. The program draws the ball once, at the start, using trigonometry, because there was no paint program to draw it in. Three tricks then make it move.
The spin is colour cycling. The screen uses 320 × 200 pixels with five bitplanes, enough for 32 colours, though only seven different colours appear. Each red or white check on the ball is drawn as six narrow slices, and each slice uses a different colour register. At the start, the registers for some slices hold red and the rest hold white, so the checks look solid. Every frame, the program moves the red and white values along the registers by one step, so the stripes of colour slide across the ball while the pixels stay where they are. Turning the ball costs the processor 28 bytes of register writes per frame. One register always holds a reddish white, placed on the border between red and white, which reads as motion blur. After 14 frames the registers are back where they started.
The bounce is scrolling. The ball never moves in memory either. The program moves the window through which the display hardware reads the screen, the operating system’s viewport, up, down, left and right, so the ball appears to travel. The playfield is 336 pixels wide, wider than the screen, and the memory around the bitplanes is kept clear so that no stray data comes into view as the window moves.
The grid stays still. The background lattice lives in the fifth bitplane. Because the ball’s colours are the same in both halves of the 32-colour palette, the ball hides the grid wherever it passes, and its shadow darkens the grid instead of erasing it. To hold the grid still while the window moves, the program moves the fifth bitplane’s starting address in the opposite direction. A bitplane pointer can only move in steps of 16 pixels, so Luck and Mical stored 16 copies of the grid, each shifted one pixel further across, and picked the right one for each frame.
The boom is a sample. The sound is Bob Pariseau hitting a closed aluminium garage door with a foam baseball bat, recorded on an Apple II inside the garage. The demo plays about 25,000 eight-bit samples through Paula, the Amiga’s sound chip. For a floor bounce it plays them at about 14,000 a second, giving a deep boom of about a second and a half, through a left and a right channel at once. For a wall bounce it plays them faster, at 22,375 a second, at about two-thirds volume and through the speaker on that side only. One recording makes both sounds.
The logo that wasn’t
Commodore launched the Amiga 1000 at New York’s Lincoln Center on 23 July 1985, and near the end of the presentation Bob Pariseau brought out an “old standby”, the Boing demo, now rewritten to run under the Amiga’s operating system. It had become a demonstration of multitasking as well: the ball bounced and boomed in one screen while other programs ran in another. Maher writes that the ball “very nearly became the machine’s official logo”, and that it appeared on pre-release Amiga hardware, including the external floppy drive in Commodore’s video of the launch. Shortly before the debut Commodore chose a rainbow-coloured tick instead, a mark closer to the logos of its 8-bit machines.
The ball stayed with the machine anyway. Dale Luck named his own company after it: in 1990 Antic’s Amiga Plus reviewed “the Boing Mouse, $125… from Dale Luck’s Boing! company”. After Commodore’s collapse in 1994, the Amiga software company Intangible Assets Manufacturing gave away metal Boing Ball logos with its orders, and in 1998 CU Amiga asked whether products endorsed by Amiga International were “worthy of the Boing Ball logo”. The next year an Amiga Format reader asked which of “the two existing logos, the Boing Ball and the tick” other readers preferred. In February 2019, according to court papers reported by the Amiga news site amiga-news.de, the “Boing Ball” design mark was among the Amiga trademarks that Amiga Inc. transferred to a company owned by Mike Battilana of Cloanto.
Bouncing balls were already an old idea in 1984. Maher points out that a bouncing-ball program ran on MIT’s Whirlwind computer around 1950, and that MOS Technology showed bouncing balls when demonstrating the video chip that went into the Commodore VIC-20. The Boing demo was remembered because of the size of its ball, the smoothness of its motion and its sound, on a computer that could do all three at once.
Sources
- Jimmy Maher, The Future Was Here: The Commodore Amiga (MIT Press, 2012), chapter 2, “Boing”, pp. 11–37: Winter CES in Las Vegas; the Whirlwind and VIC demonstrations; Amiga’s finances before CES; the Lorraine prototype, the Sage IV and the remote terminal; the all-night session; Luck’s recollection of the January version; Summer CES and Commodore’s purchase; the Lincoln Center launch and the Boing ball as near-logo; the reconstruction of the demo (resolution and bitplanes, the colour registers and 14-frame cycle, the viewport scrolling, the 16 background bitplanes, the garage-door sample and its playback rates, channels and volumes).
- BYTE, August 1985, p. 81: the prototypes shown to journalists in early 1984 and the ball whose sound moved between the speakers.
- Amazing Computing vol. 5 no. 5, May 1990, p. 54: type-in game dedicated “to the authors of the original Boing! demo: Sam Dicker, Dale Luck, and =RJ= Mical”.
- Antic’s Amiga Plus vol. 1 no. 6, February 1990, p. 73: Steve Carter’s review of the Boing Mouse from Dale Luck’s company.
- Amiga Computing issue 110, March 1997: news (Intangible Assets Manufacturing’s metal Boing Ball and tick logos).
- CU Amiga, June 1998: contents page (Amiga International endorsement and the Boing Ball logo).
- Amiga Format issue 122, April 1999: readers’ letters (the Boing Ball and the tick).
- amiga-news.de, “Legal dispute: Amiga Inc. transferred rights to company of Mike Battilana”, 5 February 2019 (web, secondary): the Boing Ball design mark in the transfer of 1 February 2019. https://amiga-news.de/en/news/AN-2019-02-00009-EN.html