30 Years Ago: The Rise, Fall and Survival of Ashton-Tate's dBASE

By Darryl Taft  |  Posted 2013-09-19

30 Years Ago: The Rise, Fall and Survival of Ashton-Tate's dBASE

Ashton-Tate's dBASE was one of the most popular early PC applications, as the Swiss Army Knife of databases when it first hit the market back in 1980.

Ashton-Tate began distributing the widely popular dBASE database management system (DBMS) for the CP/M operating system in 1980. Very quickly dBASE grew in popularity because of its simplicity and ease of use. It featured a database engine, query system, forms engine and a programming language—the dBASE language.

When IBM started shipping its PC in volume in 1982, Ashton-Tate responded by quickly porting dBASE to the platform and it became one of the few early professional programs for the PC, which only enhanced its popularity. In short, dBASE was another one of those killer apps that demonstrated the effectiveness of the IBM PC.

Typical of many high-tech startups of the time, Ashton-Tate started as a garage-based operation and grew to become one of the top three PC application software producers of its time, next to Lotus and WordPerfect—with Microsoft in the mix, as well.

Founded in 1980 by George Tate and Hal Lashlee, Ashton-Tate took what had been developed as a simple program to run an office football pool and parlayed it into a multimillion-dollar, multinational company. Ashton-Tate licensed the rights to a database program called Vulcan from its author Wayne Ratliff, who was an employee at the NASA's Jet Propulsion Laboratory. Vulcan ran on CP/M and was based on JPL/DIS, a public domain program at JPL. Besides CP/M and IBM/MS-DOS, dBASE was also ported to the Apple II.

The dBASE language enabled ease of development of custom apps, which made the database a favorite among not only hobbyists and power users, but also programmers, value-added resellers (VARs) and corporate app developers. VARs made whole businesses out of dBASE development and many companies adopted it as their standard PC database.

dBASE used a runtime interpreter architecture, which allowed the user to execute commands by typing them in a command line "dot prompt." Upon typing a command or function and pressing the return key, the interpreter would immediately execute or evaluate it. Moreover, program scripts ran in the interpreter, where each command and variable was evaluated at runtime. This made dBASE programs easy to write and test because programmers didn't have to first compile and link them before running them.

"dBASE was used to build multi-user business apps," said Philippe Kahn, CEO of Fullpower Technologies and former CEO and co-founder of Borland International, which acquired Ashton-Tate and dBASE in 1991. Users worked with the apps and "most never knew about dBASE. It was a corporate market," he noted.

In an article entitled "A Personal History of dBASE," Jean-Pierre Martel, wrote, "Because its programming language was so easy to learn, millions of people were dBASE programmers without knowing it. To use Randy Solton's words: 'In essence, dBASE brought programming power to the masses.' This popularity spawned a cottage industry of dBASE compilers or clones. dBXL, QuickSilver, Arago, Clipper and Force allowed developers to compile dBASE-III Plus programs into DOS exe files. FoxBase was a clone that was faster than the original."


30 Years Ago: The Rise, Fall and Survival of Ashton-Tate's dBASE

The PC version of dBASE launched as "dBASE II" to give buyers the impression that it was a second edition product that had any initial bugs ironed out. In 1983, Ashton-Tate released dBASE II RunTime, which allowed developers to write dBASE applications and then distribute them to customers without their needing to purchase the "full" version of dBASE. This was a hit with VARs.

Then, in May 1984, the company shipped dBASE III, a major upgrade of the program, which was written in C rather than in assembly language like previous versions. This enabled porting to other platforms such as Unix and VMS, but it brought with it the issue of legacy code, which becomes problematic for any popular system if not handled correctly.

In 1986, Ashton-Tate delivered dBASE III+, which featured character-based menus for ease of use. The product did well, and the company's sales reached $300 million in 1987.

However, some of the ease of use and simplicity of dBASE came back to bite Ashton-Tate because users began demanding more complex applications, which required enhanced programming features and better performance. Thus, competitors came in to fill the void with compilers and clones of the popular database that included such features as user-defined functions. These competitors included Fox Software's FoxBase and Nantucket's Clipper.

Meanwhile, Ashton-Tate released dBASE IV in late 1988. However, the product was late and did not include a compiler as the company had promised. It also was slow and buggy, which disappointed customers. The product received bad reviews, and sales began to drop as customers moved to alternatives. Ashton-Tate, which held 63 percent of the PC database market in 1988, saw its share drop to 43 percent in 1989; the result was layoffs as the company.

dBASE IV version 1.1 did not ship until July 1990, but by then it was too late. Ashton-Tate was looking for a suitor. The company had held merger discussions with Lotus and had entertained deals with other prominent software companies of the time, including Cullinet, Computer Associates, Informix, Symantec and Microsoft.

In the end, Ashton-Tate went to Borland for $439 million in September 1991, which was ironic because Borland already had a popular database, Paradox, which competed with dBASE. However, Borland's plan was to position dBASE as its high-end database and Paradox as its end-user database. But the pricing for the two products didn't reflect that positioning as Paradox carried a high price tag.

"dBASE was the leading installed database solution on PCs, and Borland's Paradox the technology leader," Kahn told eWEEK. "Borland had the best technology, but we couldn't penetrate some of the database-driven corporate applications."

Reviews of Paradox, which Borland acquired through its 1987 buyout of Ansa Software, were very positive, particularly among programmer types. But, as Kahn said, Borland couldn't penetrate the corporate market deeply enough to make an impact.


30 Years Ago: The Rise, Fall and Survival of Ashton-Tate's dBASE

However, "Ashton Tate was put on the auction block, and the investment bankers made a very convincing pitch that if we added installed market share to Borland's superior technology, we could become a leader for enterprise solutions," Kahn said.

Moreover, the bankers argued that if someone else like Microsoft or Lotus acquired Ashton-Tate, it would make things very difficult for Borland, he said.

"There was a lot of debate within Borland as to whether to acquire dBASE," Kahn said. "I was ambivalent. I always believe in technology and innovation first. This is the one time where I let the bankers talk us into a different path, and that was a big mistake. dBASE was legacy and we should, instead of buying dBASE, have invested in our own innovative tools.

Kahn said the dBASE acquisition became "my 'reality show of business school.' We learn the most from our mistakes."

Asked what the acquisition of Ashton-Tate did, in fact, bring to Borland, Kahn replied, "Headaches. The moment that we took responsibility for the dBASE legacy we had nothing but headaches, and we almost instantly regretted acquiring dBASE. It's a good lesson that focusing on innovation and invention are the best paths in technology. That's why ever since, I focus on developing unique, innovative IP."

Borland's goal was to take dBASE into the Windows era. But Microsoft had similar plans for its database technology. Microsoft introduced Access, the database component in Microsoft Office, in late 1992 and took over the Windows database market. Also in 1992, Microsoft bought Fox Software and its popular FoxPro database.

Prior to the introduction of Access, Borland, with Paradox and dBASE, and Fox, with FoxPro, dominated the desktop database market. However, Microsoft Access was the first major database program for Windows. With Microsoft's purchase of FoxPro and the incorporation of Fox's Rushmore query optimization routines into Access, the platform quickly became the dominant database for Windows—marginalizing the competition that failed to transition from the MS-DOS world.

dBASE still exists today and is marketed by dBASE LLC, where the latest version is dBASE PLUS 8. dBASE has evolved into a modern object-oriented language that runs on 32-bit Windows. It can be used to build a wide variety of applications, including Web apps hosted on a Windows server, Windows client applications and middleware applications. The new dBASE also can access most modern database engines via ODBC drivers.

Meanwhile, in another dBASE-related irony, Ashton-Tate and dBASE indirectly helped spawn the open-source movement when companies began to clone the dBASE components and provide third-party products. Ashton-Tate responded with threats and lawsuits.

In fact, at an industry conference, Ed Esber, CEO of the company stood up and challenged cloners with lawsuits, prompting a debate about the ownership of computer software and chants of "innovation not litigation." Yet, when Ashton-Tate's copyright infringement lawsuit against Fox Software went to court in 1990, the case was thrown out when the court found that dBASE was based on Vulcan, the public domain program created years ago at Wayne Ratliff's former employer, JPL.


Rocket Fuel