The Unexpected Final Chapter
Sometimes the most fitting endings arrive with a touch of irony. A nostalgic blog post mourning the slow death of dBase may have delivered the final blow to what remained of its online presence. The store at store.dbase.com was still operational when a piece titled “dBase: 1979-2026” appeared on a site called Delphi Nightmares. Within hours of that story being shared on Hacker News, the store vanished. This is the story of how a tool born at NASA helped define personal computing, then watched its own dbase decline unfold as rivals, missed opportunities, and the open-source movement reshaped the industry around it.
The Genesis at Jet Propulsion Laboratory
The seeds of dBase were planted far from the commercial software world. They sprouted inside the Jet Propulsion Laboratory in Pasadena, California, where engineers needed a way to manage data on the lab’s massive Univac 1108 computers. A tool called JPLDIS — short for Jet Propulsion Laboratory Data Information System — became the foundation. It started as a FORTRAN rewrite of an earlier program called Tymshare RETRIEVE. Two programmers, Jack Hatfield and Jeb Long, did the heavy lifting. JPLDIS gave scientists a way to query and organize data without writing complex machine code. It was practical, purpose-built, and quietly effective.
Wayne Ratliff Takes the Reins
C. Wayne Ratliff, a colleague at JPL, saw something more in JPLDIS. He recognized that the relational database concept could serve a much wider audience. In his spare time, Ratliff rewrote the entire tool in Intel 8080 assembly language so it could run on his personal IMSAI 8080 microcomputer under an operating system called PTSDOS. He called his creation Vulcan — a nod to his fondness for Star Trek. Ratliff placed a small advertisement in BYTE Magazine offering Vulcan for 50 dollars. The response was underwhelming. At that price and with no marketing muscle, the product barely registered in the bustling hobbyist market of the late 1970s.
Ed Tate and the Birth of Ashton-Tate
The turning point came when a sharp-eyed entrepreneur named Ed Tate entered the picture. Tate hired Ratliff and secured a license for Vulcan. He then founded a company called Ashton-Tate — a name invented purely for its memorable ring. There was no real Ashton involved until later, when Tate bought a parrot and named it Ashton to serve as the company mascot. The product was rebranded as dBASE II, deliberately skipping version one to imply maturity and stability. The price jumped dramatically from 50 dollars to nearly 800 dollars. Shrink-wrapped boxes of dBASE II began appearing on store shelves for the CP/M operating system.
The Explosive Rise in the 1980s
By 1982, dBASE II was selling at an astonishing pace. John Walker, co-founder of Autodesk, noted at the time that the software was “selling like hotcakes at $800 a pop.” That same year, a version for the IBM PC arrived, making dBase one of the first serious business applications available for the new platform. For millions of users, dBase became synonymous with the idea of a personal computer database. Small businesses, accountants, inventory managers, and even school administrators adopted it. The dot-prompt command line became a familiar sight in offices across the country.
dBASE III and III+ Refine the Formula
In 1984, Ashton-Tate released dBASE III, which offered better performance and larger file capacities. It sold even better than its predecessor. Then came dBASE III+ in 1986, which introduced a menu-driven user interface alongside the traditional command line. This made the software accessible to users who found the dot prompt intimidating. The product line was at its peak. Ashton-Tate dominated the database market with an estimated 70 to 80 percent share. Competing products existed, but none threatened dBase’s crown. The programming language embedded in dBase — a procedural language with its own syntax for data manipulation — became a de facto standard.
The Critical Misstep: dBASE IV and the Missing Compiler
In 1988, Ashton-Tate launched dBASE IV with great fanfare. The software promised many improvements, but one feature was conspicuously absent. The company had pledged to include a compiler that would translate dBase code into standalone executable programs. That compiler never materialized in the initial release. This failure created a crack in the foundation of Ashton-Tate’s dominance, and competitors rushed to exploit it. The dbase decline can be traced in large part to this single broken promise. Users who wanted to distribute applications without requiring recipients to own a full copy of dBase suddenly had nowhere to turn within the official product line.
Why the Compiler Mattered
For power users and independent developers, a compiler was not a luxury. It was a necessity. Without one, every application built in dBase required the full runtime environment on each end user’s machine. That added cost and complexity. It also made it nearly impossible to protect source code, since the.prg files remained readable by anyone who had access. A compiled executable would have solved all of these problems. When Ashton-Tate failed to deliver, developers began looking elsewhere.
Rivals Seize the Opening
Nantucket Corporation had already released a product called Clipper in 1985. Clipper compiled dBase code into standalone.exe files. Because it lacked the interactive command-line interface of dBase itself, it avoided legal challenges from Ashton-Tate. Clipper took off among developers who needed to ship commercial applications. By 1992, Computer Associates had acquired Nantucket, giving Clipper a powerful corporate parent.
Fox Software and FoxPro
Another formidable rival emerged from Fox Software. Their product, initially called FoxBase and later FoxPro, offered dramatically better performance than dBase. FoxPro could run the same dBase code several times faster, and it included a compiler from the start. Even C. Wayne Ratliff himself was impressed, noting publicly that FoxPro outperformed dBase on nearly every benchmark. Microsoft recognized the value and acquired Fox Software in 1992. FoxPro eventually evolved into Visual FoxPro, which remained in the Microsoft ecosystem until its retirement in 2007.
Other Competitors Join the Fray
The list of dBase competitors continued to grow. There was WordTech’s Quicksilver, which also compiled dBase code. Alpha Software released Alpha Five, which targeted a less technical audience. And a host of smaller players offered their own takes on the dBase formula. Each competitor ate away at Ashton-Tate’s market share, and the dbase decline accelerated.
Standardization and the Birth of xBase
As dBase’s programming language spread across competing products, a peculiar thing happened. The language itself began to escape the control of its original creator. Developers wrote code that could run on multiple platforms — dBase, Clipper, FoxPro, and others — with only minor adjustments. This interoperability made the ecosystem stronger, but it also made the original product less essential. Trademark law prevented competitors from calling their dialects “dBase,” so the community adopted the term “xBase” as an umbrella name for all implementations of the language.
The Ashton-Tate Acquisition by Borland
Ashton-Tate struggled through the late 1980s and early 1990s. Revenue declined, and the company failed to ship a compelling successor to dBASE IV. In 1991, Borland International acquired Ashton-Tate for approximately 440 million dollars. Borland owned the dBase trademark and continued to release versions, including dBASE 5.0 and later dBASE 7.0. But the magic was gone. The market had moved on to client-server databases like Oracle, Sybase, and Microsoft SQL Server. Desktop databases seemed quaint compared to the power and scale of the new generation of database systems.
The Open-Source Afterlife
Even as the commercial dBase product faded, the xBase language refused to die. Developers began creating open-source implementations that could read and write dBase file formats. The most notable of these are Harbour and its fork xHarbour. Both projects provide a compiler for the xBase language, allowing developers to build modern applications that still use the familiar syntax and data structures. The Harbour GitHub repository has seen active commits as recently as this year, and xHarbour shows signs of life as well.
What Harbour and xHarbour Offer Today
Harbour compiles xBase code into C, which a C compiler then turns into machine code. The result is a native executable that runs on Windows, Linux, macOS, and other platforms. Harbour supports many of the classic dBase commands and functions, plus extensions for modern programming concepts like object-oriented programming, threads, and network communication. xHarbour, which forked from Harbour around 2004, offers similar capabilities with a different set of trade-offs. Both projects maintain backward compatibility with dBase file formats, meaning that databases created in the 1980s can still be read and updated today.
The WAMPUM Legacy
Even before the open-source movement took hold, a DOS shareware application called WAMPUM kept the dBase torch burning. Written by an independent developer, WAMPUM could read and write dBase files and offered a scripting environment of its own. It never achieved mainstream success, but it demonstrated that the xBase ecosystem had enough vitality to sustain independent projects. Copies of WAMPUM still circulate in retro-computing circles today.
The Final Irony: A Eulogy That Knocked the Site Offline
For years, the dBase website at dbase.com remained online as a ghost of the product’s former glory. The online store at store.dbase.com still accepted orders. Then, in early 2025, a blogger writing for Delphi Nightmares published a piece titled “dBase: 1979-2026,” effectively writing an obituary for the product’s online presence. The post was shared on Hacker News, and the traffic spike may have been the final straw. The store went offline. The website that had survived for decades through neglect and inertia finally succumbed. The timing felt poetic. A eulogy caused the death it described.
You may also enjoy reading: Court grants Apple’s request to seek Samsung docs.
Lessons from the dBase Decline
The story of dBase holds lessons that resonate far beyond the world of databases. Every software vendor faces the same fundamental challenge: evolve or become irrelevant. Ashton-Tate had a dominant position, a loyal user base, and a powerful brand. But the company failed to deliver on its promises, failed to anticipate the shift toward client-server computing, and failed to understand that its own programming language had become more valuable than the product that hosted it. The dbase decline is a cautionary tale about complacency, broken commitments, and the risk of taking a user community for granted.
The Compiler Mistake Revisited
If there is a single decision that sealed dBase’s fate, it was the decision to ship dBASE IV without the promised compiler. That choice handed the market to Clipper and FoxPro on a silver platter. It taught developers that they could not count on Ashton-Tate to deliver what they needed. Trust, once broken, is extraordinarily difficult to rebuild. By the time Borland acquired Ashton-Tate and eventually delivered a compiler, the damage was done.
The Danger of Proprietary Lock-In That Backfires
Ashton-Tate tried to protect its ecosystem through proprietary file formats and a closed development model. But the strategy backfired. When the company failed to meet user needs, the community found ways around the barriers. Competitors reverse-engineered the file format, and the programming language escaped into the wild. Once that happened, dBase lost its competitive advantage. The very features that Ashton-Tate thought would lock users in actually accelerated the exodus.
What the dBase Decline Means for Modern Developers
Today’s developers work with databases like PostgreSQL, MySQL, MongoDB, and SQLite. The world of desktop databases with proprietary programming languages seems distant. Yet the patterns repeat. Modern platforms and frameworks also face the pressure to evolve, keep promises, and maintain trust. The story of dBase reminds us that no platform is too big to fail. Market dominance does not guarantee survival. And the open-source ecosystem has a way of preserving and extending the useful parts of any technology, long after the original commercial product disappears.
For Those Maintaining Legacy Systems
Believe it or not, dBase applications still run in some organizations. Small businesses, government agencies, and educational institutions sometimes rely on systems built decades ago. If you find yourself maintaining such a system, the options are clearer than they might seem. You can continue running the original dBase software under DOSBox or a virtual machine. You can migrate the data to a modern database using tools that read.dbf files. Or you can rewrite the application logic in a modern language while preserving the data. Harbour and xHarbour offer a fourth path: keep the xBase code but compile it into native executables for modern operating systems.
The xBase Language Lives On
Despite the disappearance of the dBase website and the end of dBase LLC, the xBase language continues to see active development. The Harbour project has contributors from around the world, and the language has been extended with features that would surprise anyone who remembers the dot prompt of the 1980s. Modern xBase supports object-oriented programming, garbage collection, multithreading, TCP/IP networking, and bindings to popular C libraries. It can generate graphical user interfaces using GTK, Qt, or native Windows APIs. It can access SQL databases through ODBC. And it can still read and write.dbf files with perfect fidelity.
Why xBase Still Matters
For a certain generation of programmers, xBase is not a historical curiosity. It is a productive tool that solves real problems today. The language is relatively easy to learn, especially for people who already understand procedural programming. It handles data manipulation tasks concisely. And because it compiles to native code, it offers performance that rivals C in many data-intensive workflows. The hobbyist and retro-computing communities have embraced xBase as a way to preserve and extend the software heritage of the microcomputer era.
Fragility of Digital History
The disappearance of the dBase website highlights a broader concern about digital preservation. How much of our software history will survive the next few decades? Websites go offline. Source code disappears from repositories. Documentation rots when the domain name expires. The story of dBase is better documented than most, thanks to the efforts of historians, bloggers, and open-source contributors. But countless other products have vanished with barely a trace. The dbase decline and the subsequent loss of its online presence serve as a reminder that digital history is fragile. Preserving it requires intentional effort.
What Can Be Done
Organizations like the Internet Archive and the Software Preservation Network work to capture and store software artifacts. Individuals can contribute by uploading old software, manuals, and source code to archival repositories. If you have old dBase disks or documentation sitting in a box, consider donating them to a preservation project. Every piece of the puzzle helps future historians understand how we got from there to here.
The Bigger Picture: From Desktop to Cloud
The dbase decline mirrors a much larger shift in the database industry. In the 1980s, a database was something you installed on your personal computer. It served one user at a time, or perhaps a small network of users. Today, databases run in data centers, scale horizontally, and serve millions of users simultaneously. The shift from desktop to client-server to cloud has left many once-dominant products behind. dBase was not alone in its fate. WordPerfect, Lotus 1-2-3, and many other desktop applications suffered similar declines. The lesson is clear: technology moves forward, and products that do not move with it get left behind.
The Community Remembers
Perhaps the most heartening aspect of the dBase story is the community that still surrounds it. Forums, mailing lists, and GitHub repositories keep the conversation going. Retired developers share tips and tricks from decades past. Newcomers discover xBase and wonder why they never heard of it before. The shared memory of the dot prompt, the.dbf file format, and the peculiar syntax of the dBase programming language binds people across generations. That community will outlast any company or website.
