Martin L. Rinehart

A Fast History of the DBF

In the 1970s, NASA's Jeb Long invented a data-handling language and implemented it on a mainframe computer at the Jet Propulsion Laboratory (JPL) in Pasadena. (Rocket science has more sex appeal, but space exploration involves keeping track of lots of nuts and bolts, too.) Long's language, known as JPLDIS ("JIP-uhl-dis"—Jet Propulsion Laboratory's Display Information System) made it easy to define and implement tables. It enabled you to add, edit, and delete records in these tables. And it let you perform the basic relational operations, such as joining tables.

Another programmer at JPL, Wayne Ratliff, had an early, 8-bit microcomputer at home as a hobby. He decided to implement Long's fourth-generation language (4GL) on his microcomputer. The first application was to keep track of the office football pool.

Wayne worked in Z-80 assembler. (The Z-80 was an 8-bit microprocessor that dominated the early, hobbyist period of the development of personal computers.) He managed to code the database work and an interpreter for the 4GL, and to still provide data space within his 48K RAM space. (The early operating system, CP/M, took less than 4K of this space.)

Wayne began marketing his program to other hobbyists. He named it Vulcan. It came to the attention of Hal Lashlee and George Tate, who were running an early software distribution company out of Tate's garage. They acquired marketing rights and renamed the product dBASE II.

George partnered with a marketing professional, Hal Pawluk, in promoting the product. (Lashlee, an accountant, didn't get involved in day-to-day operations.) Through a combination of brilliant marketing and the simple fact that dBASE II was a superior product in its day, this became the leading database-management software for CP/M-based computers. They named their company Ashton-Tate.

Would they have succeeded selling "dBASE I" from "Lashlee-Tate"? They might have, because the product was better than its competitors. Certainly success came more rapidly and market dominance was more complete because of Ashton-Tate's early marketing orientation.

The introduction of the PC in 1981 and then of the PC-XT in 1983 (the latter featured a whopping 10MB hard disk, the first from a major manufacturer) helped make the personal computer a viable machine for small database work. While some companies stumbled (the inventors of the spreadsheet lost their market lead to PC-specific upstart Lotus Development; the dominant WordStar word-processing program was overtaken by WordPerfect), Ashton-Tate continued to dominate databases.

The dBASE data format was published by Ashton-Tate, which actively encouraged others to use it. Spreadsheets read DBF files (tables map nicely into spreadsheets). Word processors used the DBF format for mail-merge applications. Graphics programs read DBFs. (In the early days, graphing was a separate application, not part of every spreadsheet and word processor.) And, of course, competing database products accessed DBF-based data.

The dBASE product is still available from Borland, although that company has hinted that they may have released the last version of the product. The Xbase language is still used in Microsoft's FoxPro and CA's Clipper products. And the DBF format is still a widely used, de facto standard in numerous PC applications.

(Source: "Java Database Development" ISBN: 0078823560)

James Copeland

Circa 1980, a programmer for the Jet Propulsion Laborotries, by the name of Wayne Ratlief, imported a mainframe database program into the mini environment and called it Vulcan. Wayne later hooked up with a character, a salesman, named George Tate. (I've never seen any report that George Tate had any programming experience.) dBASE version 2.0 was born. The company was called Aston-Tate. (There never was anyone named Aston, it just sounded good and there never was a version 1 of dBASE.) dBASE was improved to the point that dBASE 3+ became the premire and dominate desktop industry standard. Finally, dBASE 4 version 1.0 was released. But Wayne had left the company by then and dBASE 4 was actually created by a committee. It has the distinction of being probably one of the most buggy mainstream programs ever, including the early Windows programs.

But George Tate decided that the MacIntosh was obviously the wave of the future, and the resources of Aston-Tate were directed towards porting dBASE into the Mac environment. (Question, why *didn't* the Mac become the 800 pound gorilla?) It took almost a year and a half before a bug-fix program, dBASE IV version 1.5 was released and eventually dBASE IV version 2.0. Version 2.0 was really pretty good. Fast, stable and useable.

However, in this same timeframe, a University professor (Indiana, as I remember) and his grad students, had created another program, that eventually became Foxpro. Most database application designers will tell you that Foxpro was better than dBASE (at that time). But dBASE had the corporate market tied up. dBASE was the defacto standard. (dBASE, Lotus 1-2-3 and Wordperfect were the Big 3). But because dBASE had released such a rotten program AND they had suspended fixing the release for such an extended period of time, they gave the new an innovative Foxpro a chance to gain a foothold on the ever expanding market. Bad decisions, bad timing, bad news.

Aston-Tate never recovered from a series of bad decisions, and was eventually bought out by Borland.

Foxpro was eventually bought by Microsoft, and the technology, Jet, was gradually merged into Access (which, until that point had not gone anywhere). Access is probably the industry dominate database program today. However, part of that dominance is not based on the superior Jet technology. Gradually, the ability to utilize Visual Basic to expand the useability of Access (Word and Excel, for that matter), has contributed to the dominance. Aggressive marketing plays a part.

The real gensus of Microsoft was the marketing of the Office Suite. (I'm not sure who came up with that idea.) So, for example, WordPerfect might be better than Word, but Wordperfect is really competing against Word and Access. dBASE 5, version 5.5 might be better than Access, but it's completing with Access and Excel. So, in every single instance, someone might be able to point to a better individual program, the totality of the programs in the Suite most generally provided a clear winner. And whatever else can be said about Microsoft programs, they've always tended to work well together.

Microsoft has exhibited a most uncanny expertise to correctly guage their target market. Most on this NG, might bemoan the "bloat" in Microsoft programs, and be entirely justified in that critism, but the vast majority of users don't know and don't care. And they have put their money behind that attitude. Note the "fast boot" facility. Users (customers) didn't like waiting the minute or two it takes to do a cold boot. It was irritating. So...they've reacted to that.

Have there been failures. Sure. How many people use Money rather than Quicken? (Microsoft did try to buy Intuit to the tune of 2.2 billion dollars as I remember.)

However, Bill Gates knows, that sooner or later, Microsoft is going to stumble. (It took a completely new technology, but the really big gorilla, namely IBM, banged right into a wall.) And Microsoft, itself, came very close in 1994. Will Linux break the juggernaught? Myself, I doubt it. Unless they can create a Linux 64-bit operating system that could operate on any Pentium class machine and sell it for under $100. Then, there might well be a "horserace".

Microsoft has reached the self-sustaining mass category. To beat them out now will take something now only new, but much better AND something that Microsoft cannot duplicate in a short timeframe.