The History of Computer Solitaire: From PLATO to Browsers

Short answer: Computer solitaire developed through several distinct paths rather than one “first” program. Paul Alfille created FreeCell for the networked PLATO system in the 1970s; Wes Cherry wrote Windows Solitaire while interning at Microsoft in 1988, and it shipped with Windows 3.0 in 1990; Jim Horne wrote Microsoft FreeCell after encountering the game on PLATO; and Microsoft distributed Spider Solitaire with Plus! 98 in 1998. Software then added rule enforcement, repeatable deals, statistics, undo, hints, and cross-device play.

Physical solitaire already suited computation: it has explicit states, legal transitions, hidden or visible information, and a clear completion test. Yet putting cards on a screen changed the experience. A program could shuffle instantly, prevent accidental illegal moves, remember an exact deal, animate multi-step transfers, and collect results across players. Those capabilities affected which variants spread and how their rules became standardized.

This history avoids a common trap. A documented early program is not automatically the first computer solitaire ever written. Small programs could circulate on mainframes, terminals, home computers, and private systems without surviving documentation. The goal here is to follow well-supported milestones that shaped modern play.

Contents

• What computers added to a card game
• PLATO and Paul Alfille’s FreeCell
• Windows Solitaire: 1988 to 1990
• Why mouse training and entertainment are both part of the story
• Jim Horne and Microsoft FreeCell
• Spider arrives with Microsoft Plus! 98
• From desktop programs to browser solitaire
• Paper rules, software rules, and interface policy
• A documentary timeline
• In brief
• Frequently asked questions

What computers added to a card game

A computer version can reproduce the visible mechanics of a card layout, but its deeper contribution is state management.

Instant setup and shuffling

A physical Klondike game requires shuffling and dealing 28 tableau cards in a staircase before play begins. Spider needs two decks and a ten-column arrangement. Software performs the setup consistently and can start another deal immediately.

For Paul Alfille, avoiding the need to reshuffle a deck sorted into foundations was a practical reason to program FreeCell SRC-025. Convenience was not separate from design: removing setup friction made repeated analytical play much easier.

Rule enforcement

Software can reject a same-color Klondike placement, a mixed-suit Spider group move, or an over-capacity FreeCell transfer. This reduces accidental rule violations and allows beginners to learn by interaction.

Enforcement also makes implementation differences consequential. If one program permits a move another rejects, the programs may encode different empty-column, sequence, or automation rules. The screen does not create a universal standard by itself.

Automatic consequences

Programs can turn an exposed face-down card, remove a complete Spider run, or move a safe card to a foundation. Some actions simply automate a physical step; others embody a policy choice. An aggressive auto-foundation system may reduce decisions, while optional automation preserves them.

Repeatable and shareable deals

A physical shuffle is difficult to reproduce exactly. Software can assign a number or seed to a deal, allowing the same arrangement to be replayed, studied, or offered as a daily challenge. Repeatability supports solver research and player comparison.

Statistics and competition

A central system can store wins, streaks, and tournament results. Alfille’s PLATO implementation tracked competitive streaks and supported configurable game dimensions SRC-025. That social layer was possible because many terminals reached the same computer environment.

New interface vocabulary

Dragging a card, double-clicking a foundation move, tapping the stock, pressing undo, and requesting a hint are software interactions. They sit beside the card rules rather than replacing them. This distinction becomes central to accurate guides and localization.

PLATO and Paul Alfille’s FreeCell

PLATO as a networked environment

PLATO began at the University of Illinois as a computer-assisted instruction system. It grew into a distributed environment with terminals, programming tools, messaging, and games; by the 1970s, users at multiple institutions could participate in a shared computing culture SRC-027.

That context matters. FreeCell was not merely a local card animation. It could be played on a central system that recorded performance and exposed a program to a networked community.

Alfille’s recollection of the paper game

In a 2000 interview, Paul Alfille recalled playing a related solitaire game with physical cards as a child. He did not confidently identify the book or the exact rule change that separated his version from its predecessor, and he believed he coined the name FreeCell SRC-025.

Later summaries often draw a clean line from Eight Off to Baker’s Game to FreeCell. The games are historically related, but Alfille’s own uncertainty should remain visible. The digital creation is well documented; the exact paper lineage is less certain.

Programming FreeCell in TUTOR

Alfille programmed the game in TUTOR, PLATO’s authoring language, while he was a medical student SRC-025. Hardware constraints shaped the implementation. He described a 512 × 512 plasma display and a configurable range of tableau columns and cells, rather than only the later standard eight-column, four-cell format.

The program included features that would now be recognized as product design:

• selectable column and cell counts;
• central records;
• winning streaks;
• competitive use;
• tournaments or shared challenges;
• simple detection of positions with no productive continuation SRC-025.

It was not a modern exhaustive solver. Alfille described a limited procedure for recognizing simple losing loops, constrained by the language and processor environment. That distinction is important: computer enforcement and computer proof are different capabilities.

Why PLATO FreeCell was historically important

PLATO FreeCell connected three traditions:

1. a physical open-card patience layout;
2. a programmable rules engine;
3. a networked community with persistent records.

The standard modern 8 × 4 layout was not the only form Alfille explored, but it became the dominant public version through later software. The broader PLATO implementation shows that variant configuration and competitive solitaire were present long before browser game menus.

For the card-game lineage, see the history of FreeCell.

Windows Solitaire: 1988 to 1990

Wes Cherry’s program

Reporting based on a later interview with Wes Cherry says he wrote Solitaire for Windows while interning at Microsoft in 1988 SRC-034. He described using idle time to create a game for an environment that had relatively few built-in diversions.

This is a creator account of the Windows program, not a claim that Cherry invented Klondike. He adapted an existing card game to a graphical operating system.

Release with Windows 3.0

Microsoft’s own retrospectives state that Windows Solitaire shipped with Windows 3.0 in 1990 SRC-028 SRC-029. The built-in game presented Klondike through windows, card graphics, pointer selection, and drag-and-drop interaction.

The release date is much better documented than broad claims about “the first computer solitaire.” PlaySoli therefore uses the precise formulation: Windows Solitaire shipped with Windows 3.0 in 1990.

Susan Kare’s card designs

Microsoft credits Susan Kare with the card faces used by Windows Solitaire through the versions before Windows Vista SRC-029. The graphics had to communicate suit, rank, selection, and overlap at low resolution. Their longevity shows how interface design can become part of a game’s cultural memory even when the underlying rules are centuries older.

The cascading-card reward

The familiar end animation transformed a completed foundation layout into a visual celebration. It did not alter the win condition, but it made completion unmistakable and gave the digital version a distinctive reward unavailable in ordinary physical play.

Why mouse training and entertainment are both part of the story

Microsoft describes Windows Solitaire as helping users learn mouse control and drag-and-drop interaction when those actions were unfamiliar SRC-028 SRC-029. A card game was well suited to the task:

• users pointed at discrete objects;
• clicked and held a card or sequence;
• dragged it to a visible destination;
• released it and received immediate feedback;
• repeated the interaction without a formal tutorial.

Cherry later said the practical motivation was entertainment and that the official mouse-training explanation did not capture the whole personal story SRC-034. These accounts need not conflict.

A programmer can create a game because it is enjoyable. A product team can include the same game because it teaches an interface. Users can then treat it as a break, a puzzle, or their first encounter with graphical computing. Historical motives can coexist at different organizational levels.

The careful editorial formulation is therefore:

Windows Solitaire was written as a game and was also deployed as an approachable way to practice mouse and drag-and-drop interaction SRC-028 SRC-034.

It is less accurate to reduce the program to either “only training” or “only office distraction.”

Jim Horne and Microsoft FreeCell

From PLATO to personal computers

Jim Horne has described encountering FreeCell on a PLATO system at the University of Alberta SRC-026. He wrote versions for personal computers, including an early text-oriented implementation distributed through CompuServe, before creating a graphical Windows version.

This route is historically important because it carried a game associated with a university mainframe network into the personal-computer software ecosystem.

The Windows version

Horne’s account describes building FreeCell for Windows and inserting it into Microsoft development builds, first around Windows NT work and then into the Windows 95 line SRC-026. The story includes an internal joke about the program’s credit, but the durable result was straightforward: FreeCell became a bundled Windows game familiar to users who had never seen PLATO.

The Microsoft version standardized several expectations for a wide audience:

• eight tableau columns;
• four free cells;
• four foundations;
• fully visible cards;
• numbered or repeatable deals in familiar distributions;
• graphical movement with automated multi-card transfers.

The interface made sequence movement look like a direct group rule, although the logic can be understood as a series of one-card moves through free cells and empty columns. That difference between primitive rules and animation is still important in FreeCell supermove explanations.

Creator and distributor are different roles

Alfille created the PLATO program and likely the FreeCell name. Horne wrote the Microsoft Windows implementation that spread through Windows. Neither claim requires erasing the related physical-card games that preceded the software SRC-025 SRC-026.

Spider arrives with Microsoft Plus! 98

Spider existed in print well before modern Windows distribution. A 1949 book documents a form close to modern play, and specialist research reports earlier evidence SRC-022 SRC-023. The computer milestone is separate.

Microsoft’s official release dated June 25, 1998 lists Spider Solitaire among the games included with Microsoft Plus! 98 SRC-024. That product accompanied the Windows 98 launch and helped expose the two-deck game to a much broader personal-computer audience.

The software format suited Spider particularly well:

• a two-deck, ten-column setup appears instantly;
• legal group movement can be checked automatically;
• new rows are dealt evenly across columns;
• the “no deal while a column is empty” rule can be enforced;
• complete same-suit King-to-Ace runs can be recognized and removed;
• suit-level variants can share one interface.

The 1998 release is a documented distribution milestone, not the invention date of Spider. See the history of Spider Solitaire for the 1937/1949 evidence conflict.

From desktop programs to browser solitaire

Desktop bundling made particular implementations familiar. Browser delivery changed the distribution model again: a player could open a game without installing an operating-system package, and one codebase could serve multiple locales and devices.

Client-side state and instant access

Modern browser games can keep the deal and move history in local memory, render cards responsively, and restart immediately. They can support mouse, touch, and keyboard input from the same game state.

Responsive layouts

A physical card ratio and a ten-column Spider tableau do not naturally fit every screen. Browser implementations must decide how cards overlap, how columns compress, when scrolling appears, and how touch targets remain usable. These are interface decisions with direct effects on readability but should not alter card legality.

Daily and seeded deals

A deterministic seed can reproduce a deal across devices and locales. Daily Solitaire turns that capability into a shared dated challenge. “Daily” describes identity and timing; it should not be assumed to mean solver-verified unless the product documents that process.

Localization

A digital game can keep the rules stable while presenting natural terminology in each language. PlaySoli uses established local names for areas such as the stock, waste, tableau, and foundations so players do not have to decode literal English labels.

Accessibility and input diversity

Software can offer card labels for assistive technology, high-contrast display, reduced motion, keyboard navigation, larger touch targets, and text explanations of rejected moves. These features extend access beyond what a purely visual drag interface provides.

Data without unsupported conclusions

Programs can count games, moves, time, undo use, and completions. Those measurements are useful only with defined denominators and privacy-respecting collection. A completion percentage from self-selected sessions is not automatically the mathematical winnability of the deal population.

Paper rules, software rules, and interface policy

A reliable digital guide separates three layers.

Core card rules

These define layout, legal movement, stock behavior, and winning condition. Examples include alternating colors in Klondike, same-suit group movement in Spider, and one card per free cell.

Mechanical automation

These actions reproduce a necessary physical step:

• shuffling and dealing;
• turning an exposed face-down card;
• removing a complete Spider run;
• returning waste cards for an allowed redeal.

Interface policy

These features may vary without changing the nominal game family:

• unlimited or limited undo;
• hints;
• auto-finish;
• animation speed;
• whether a double-click sends a card to a foundation;
• whether legal destinations are highlighted;
• how a FreeCell supermove is animated;
• whether a move counter counts animation as one action or several.

Some policies can affect measured play. An auto-finish that chooses only forced moves may be neutral to solvability, while an auto-foundation system that removes useful cards could change available choices. Research and product copy should state the behavior rather than assume it.

PlaySoli’s current exact rules are documented in Solitaire Rules Explained.

A documentary timeline

Period	Documented milestone	Editorial boundary
1960s–1970s	PLATO develops from an Illinois educational system into a distributed environment with games and communication SRC-027.	Not a claim that every early card program is known.
Mid-1970s	Paul Alfille creates FreeCell for PLATO while a medical student, using TUTOR SRC-025.	The exact paper predecessor remains uncertain.
1988	Wes Cherry writes the Windows Solitaire program while a Microsoft intern, according to later interview reporting SRC-034.	Cherry created the Windows adaptation, not Klondike itself.
1990	Windows Solitaire ships with Windows 3.0 SRC-028 SRC-029.	A mass-distribution milestone, not proven first computer solitaire.
Early Windows era	Jim Horne writes graphical Microsoft FreeCell after encountering PLATO FreeCell SRC-026.	The Windows path does not erase Alfille or paper relatives.
1998-06-25	Microsoft announces Plus! 98 with Spider Solitaire SRC-024.	Distribution date, not Spider invention date.
Browser era	Solitaire gains install-free access, responsive layouts, seeded deals, localization, and cross-device input.	Features and exact rules remain implementation-specific.

In brief

• Computer solitaire has several lineages; no universal “first program” is asserted here.
• PLATO provided a networked environment in which Paul Alfille created and named FreeCell, added configuration, and tracked competition.
• Wes Cherry wrote Windows Solitaire as an intern in 1988; it shipped with Windows 3.0 in 1990.
• Microsoft describes the game as helping users practice mouse and drag-and-drop control; Cherry also emphasized entertainment.
• Susan Kare created the classic Windows card faces used for many versions.
• Jim Horne wrote the Microsoft FreeCell implementation after encountering the game on PLATO.
• Microsoft officially shipped Spider Solitaire with Plus! 98 on June 25, 1998.
• Browsers added immediate access, shared deals, responsive interfaces, localization, and new accessibility possibilities.

Common historical mistakes

Calling Windows Solitaire the first computer solitaire

It was a major mass-distribution milestone. Earlier mainframe and personal-computer card programs existed, and the complete record is difficult to reconstruct.

Saying Microsoft invented Solitaire, FreeCell, or Spider

Microsoft distributed influential software versions. Klondike and Spider predate those versions; Alfille’s PLATO FreeCell predates Horne’s Windows program.

Treating mouse training as the only motive

Official product purpose and a programmer’s personal motive can differ. The sources support both interface training and entertainment.

Treating automatic actions as timeless rules

A physical-card player manually flips cards and removes runs. A program automates those steps and may add optional policies such as hints or auto-finish.

Using a software release date as a game invention date

Windows 3.0 dates one implementation of Klondike; Plus! 98 dates one distribution of Spider. Neither dates the underlying card game’s creation.

Assuming data automatically answers winnability

Session statistics measure user behavior under product conditions. Mathematical solvability requires a defined ruleset, deal population, and analytical method.

Frequently asked questions

What was the first computer solitaire game?

The exact first is not established in this guide. Programs could exist on poorly documented systems. PLATO FreeCell is an important well-documented early networked implementation, while Windows Solitaire is a landmark in mass distribution.

Who created computer FreeCell?

Paul Alfille created the FreeCell program for PLATO and believed he coined the name SRC-025. Jim Horne later wrote the Microsoft Windows version after learning the game through PLATO SRC-026.

When was Windows Solitaire released?

It shipped with Windows 3.0 in 1990 SRC-028 SRC-029. Wes Cherry had written the program while interning at Microsoft in 1988, according to later interview reporting SRC-034.

Was Windows Solitaire really designed to teach the mouse?

Microsoft states that it helped users learn mouse and drag-and-drop interaction SRC-028 SRC-029. Cherry’s account says he wrote it for fun during internship downtime. Both can be true: creation motive and deployment value are different questions.

Who designed the classic Windows cards?

Microsoft credits Susan Kare with the card designs used from the early Windows version through the pre-Vista era SRC-029.

When did Microsoft release Spider Solitaire?

Microsoft officially listed Spider Solitaire in Plus! 98, announced on June 25, 1998 SRC-024. The card game itself is older.

Did computers change the rules of solitaire?

They standardized particular implementations and added automation. Some core rules came directly from card play; other conventions—undo, hints, automatic foundation policies, and supermove animation—are software-specific.

Why was FreeCell a good computer game?

All cards are visible, the legal moves are explicit, and software can avoid tedious shuffling, track repeatable deals, calculate group-transfer capacity, and preserve records. Alfille specifically recalled programming it partly to avoid reshuffling sorted physical cards SRC-025.

What is the difference between a numbered deal and a daily deal?

A numbered deal has a stable identifier within a generator or collection. A daily deal assigns an arrangement to a date. Either can be reproducible; neither is necessarily guaranteed solvable unless the product says so.

What did browsers add beyond desktop solitaire?

They reduced installation friction, enabled responsive and touch interfaces, simplified localization and updates, and made shared seeded or daily deals accessible across devices.

Related PlaySoli guides

• The broader history of card solitaire
• FreeCell from related card games to PLATO and Windows
• Klondike before and after Windows
• Spider before Microsoft Plus! 98
• How software and solvers measure winnability
• Core rules versus interface behavior

Sources used

• SRC-001 PlaySoli implementation and editorial specification.
• SRC-010 David Parlett — card-solitaire historical context.
• SRC-022 Morehead and Mott-Smith — 1949 Spider documentation.
• SRC-023 Michael Keller — pre-1949 Spider evidence.
• SRC-024 Microsoft — Plus! 98 availability announcement.
• SRC-025 Interview with Paul Alfille.
• SRC-026 Interview with Jim Horne.
• SRC-027 University of Illinois — PLATO institutional history.
• SRC-028 Microsoft/Xbox — 30 years of Microsoft Solitaire.
• SRC-029 Microsoft/Xbox — Windows Solitaire retrospective and Susan Kare credit.
• SRC-033 Michael Keller — FreeCell technical history.
• SRC-034 The Verge — reporting based on an interview with Wes Cherry.

Material checked: 2026-07-17.

Disputed or unverified facts: No absolute first computer-solitaire program is claimed. Alfille’s exact paper predecessor remains uncertain. The Windows mouse-training role and Cherry’s entertainment motive are presented as compatible accounts from different perspectives. Windows and Plus! release dates are not treated as invention dates for the underlying card games.

Editorial responsibility: PlaySoli Editorial Team.