The Canon Cat — Manual and Documentation

Archive Notice: This page is part of the Jef Raskin historical archive, preserved for its academic and historical significance.

The Canon Cat was a dedicated word processing computer released by Canon in 1987, designed by Jef Raskin. It was his most complete commercial implementation of the interface principles he had been developing since the Macintosh project — and in several respects a more faithful realization of those principles than the Macintosh had been.

The Canon Cat was sold for approximately $1500 and discontinued after six months on the market, reportedly due to internal Canon corporate decisions rather than any failure of the product itself. It has since become a significant object in the history of alternative computing — studied by interface designers and historians as evidence that a radically different approach to personal computing was commercially viable.

What the Canon Cat Was

The Cat was not a general-purpose computer. It was designed to do one thing — handle text — and to do it extremely well. This was intentional. Raskin had long argued that the “personal computer” paradigm, in which one device attempts to handle every computing task, was a design mistake that imposed unnecessary complexity on users who needed only a subset of those capabilities.

Most users, most of the time, needed to write. Letters, memos, reports, notes. The Cat was designed for those users. It did not run spreadsheets or databases or games. It ran your document, and it was always on, always ready, always in your document when you turned on the screen.

The experience of using the Cat was described by users as dramatically more immediate than using a PC or Macintosh. You pressed a key and a letter appeared. There was no boot sequence, no startup delay, no application to launch. The computer was a dedicated tool, like a typewriter that remembered everything you had ever typed.

Hardware

The Canon Cat was built around a Motorola 68000 processor running at approximately 5 MHz. It had 256 KB of RAM and a built-in 9-inch monochrome CRT display with a resolution of 640 × 400 pixels. Storage was provided by 3.5-inch floppy disks in a proprietary format.

The Cat included a built-in 1200/300 baud modem for telecommunications — a forward-looking feature for 1987. The keyboard was full-size and high quality, reflecting Raskin’s conviction that text entry was the primary interface and deserved appropriate hardware.

The system software was written in Forth, chosen for its efficiency and real-time characteristics. The Forth environment allowed rapid modification of the system’s behavior and was used to implement the Cat’s unusual interface model.

The defining feature of the Cat was its LEAP navigation system. Rather than menus, dialog boxes, file browsers, or toolbar buttons, the Cat offered a single, consistent mechanism for finding and acting on anything in the document: LEAP.

The LEAP keys — two keys on the keyboard, one for searching forward and one for searching backward — allowed instant text search. To navigate to any word or phrase in your document, you held the LEAP key and began typing. The cursor jumped to the nearest occurrence of those characters as you typed. Release the key and the cursor was positioned. Hold again and continue typing to refine.

Commands worked the same way. To execute a command, you held LEAP and typed the command’s name. The system searched forward through the document and found the command — because commands were simply text in the document, accessible via the same search mechanism as any other text. There was no separate command space. There was no application menu. Everything was in one place, found one way.

This was not merely clever — it was a coherent design principle. The LEAP mechanism was one thing, learned once, applied everywhere. Users who learned to navigate with LEAP had learned the entire interface. There were no additional paradigms to master.

Raskin later described LEAP as the most successful implementation of his interface principles he ever achieved. Users who learned it found it fast and comfortable. The challenge was that it was unfamiliar, and users who had extensive experience with conventional interfaces sometimes found the absence of menus more disorienting than the presence of LEAP was helpful.

The Document Space

The Cat presented a single, continuous document space. There was no “file system” in the conventional sense. All of your documents existed in a single persistent stream. You moved through that stream using LEAP and time navigation. Documents were separated by markers in the stream, not by the file system hierarchy.

This was a radical departure from the desktop model, which organizes documents as discrete files within hierarchical folders. Raskin’s argument against the folder model was that it externalized document organization into an abstract structure that had little to do with how people actually thought about their documents. People remember content. They remember approximately when they wrote something. They remember what it was about. They do not reliably remember where they filed it.

The Cat’s single-stream model allowed navigation by content — you found documents by searching for words you remembered from them — and by recency — the document you last edited was where the cursor last was. These were more natural retrieval cues than folder location.

The single-stream model had practical limitations, particularly around sharing documents with others who used conventional computers. But within its intended use — a personal document repository — it was elegant and efficient.

The Manual

The Canon Cat’s manual was a product of the same design philosophy as the Cat itself. Raskin believed that a well-designed interface should require minimal documentation — that if a user needed to consult a manual to perform a routine operation, the interface had failed. The Cat’s manual was therefore short, organized around learning LEAP as a foundational skill, and structured to enable rather than substitute for direct experience with the machine.

The manual covered the LEAP system in detail, explaining both the navigation and command functions. It described the telecommunications capabilities, the document management conventions, and the system configuration options. It also included tutorials that encouraged users to learn by doing rather than by reading.

The technical reference portions of the Cat documentation addressed the Forth programming environment, allowing power users to customize and extend the system’s behavior. This extensibility was one of the Cat’s less-publicized features — the Forth foundation made the Cat more programmable than it appeared.

Legacy

The Canon Cat lasted six months in the market and was never updated or replaced. Raskin never received a satisfying explanation of why Canon discontinued it. His account was that internal politics — a dispute between Canon’s professional division, which had developed the product, and its consumer division, which was responsible for selling it — led to the discontinuation. Sales had been proceeding.

The Cat became influential in its discontinuation in a way it might not have been in commercial success. Interface designers, historians of computing, and enthusiasts of alternative computing history have studied it extensively as evidence of a road not taken. Its LEAP navigation system directly informed the design of Archy. The principle that universal search could replace menus reappears in contemporary operating systems — Spotlight, Alfred, Quicksilver, the browser omnibox — though typically not as deeply integrated as in the Cat.

Functioning Canon Cats are rare. They are sought by collectors and researchers. The machine’s documentation — including this manual — is preserved here as part of the Jef Raskin Archive.

This page is part of the Jef Raskin Archive, preserving the published works and documentation of Jef Raskin, creator of the Macintosh project at Apple.

The Raskin Center