IDE Toolkit

Room 2 · The Blue Box

Borland, 1983 to 2009

The company that priced the compiler and paid for it later

Employees work at desktop computers in a dimly lit 1980s-style office

A software company at work: the shape Borland grew into, and then contracted out of.

Philippe Kahn founded Borland International in Scotts Valley, California, in 1983 with a product designed to upend the economics of professional software tools. Turbo Pascal, released in November of that year at $49.95, put a single-pass compiler, an editor and a linker inside a program that fit entirely in RAM and responded in seconds. A comparable Pascal compiler from a university or a larger vendor cost hundreds of dollars and ran as a batch job. Borland's price was not a loss leader; it was a statement of what the tooling market would eventually become.

The technical engine behind Turbo Pascal was Anders Hejlsberg, then in his early twenties, who had written the core compiler in assembly language tight enough to stay memory-resident. The product sold more than 400,000 copies in its first year, establishing Borland as a developer-tools company with a genuine mass market — something that had not existed before in quite that form. Where IBM and Microsoft priced compilers for institutions, Borland priced for individuals, and individuals responded.

A hand rests on a laptop trackpad beside glasses, with code displayed on screen

Forty years on, the same loop: type, compile, read the error, type again.

Photo: Daniil Komov / Pexels

Expansion, acquisition and the Delphi moment

Through the mid-to-late 1980s Borland expanded the Turbo brand systematically: Turbo C in 1987, Turbo Assembler and Turbo C++ following in 1989 and 1990. Each product repeated the formula — professional capability, student price, immediate response to keystrokes — and each grew the company's footprint among working developers. By the end of the decade Borland was acquiring as well as building. The 1991 purchase of Ashton-Tate brought Borland dBase, then the dominant database application on personal computers, along with a large corporate customer base and a legal entanglement: Ashton-Tate had sued Fox Software over dBASE compatibility, and Borland inherited the suit along with the product. The courts eventually ruled that the dBASE programming language could not be copyrighted, a decision that had further implications when Borland's own spreadsheet, Quattro Pro, became the subject of copyright litigation with Lotus.

The early 1990s were simultaneously Borland's peak and the beginning of its unravelling. Revenue exceeded $400 million by fiscal 1994. The company employed thousands. But the litigation costs were real, the Ashton-Tate integration was complicated, and Microsoft was quietly reorganising its developer-tools business around a forthcoming environment that would redefine the category.

Two people work at glowing desktop monitors displaying code in a dim room

One workstation, one screen, and a component library you could open and read.

Delphi arrived in February 1995 and was, for a moment, the most-discussed development environment in the Windows ecosystem. It combined a visual form designer — a canvas on which the developer could drag UI components from a palette and produce working, event-connected applications — with Object Pascal, a compiled language that produced fast native executables, and the Visual Component Library, an extensible framework that developers could subclass and redistribute. The VCL's open architecture meant that a commercial market in third-party components formed almost immediately. Delphi 1.0 sold out its initial production run and generated a level of enthusiasm in the technical press that Borland had not seen since the Turbo Pascal years.

The price war Borland could not win

Microsoft shipped Visual Basic in 1991 and Visual C++ shortly after, but Visual Studio 97 — bundling multiple environments into one shell — was the product that changed Borland's situation structurally. Microsoft had both the financial depth to sustain a price war and the operating-system relationship that gave its tools a perpetual integration advantage. When Microsoft began bundling development tools with MSDN subscriptions at prices Borland could not match, the economics that had once favoured Borland's aggressive pricing reversed entirely. Borland was now the smaller company trying to undercut the larger one, without the margins to sustain it.

The Blue BoxWhere IBM and Microsoft priced compilers for institutions, Borland priced for individuals, and individuals responded.

The mid-1990s saw Borland attempt to answer on several fronts. C++ Builder arrived in 1997, bringing the visual-form-design model of Delphi to C++ development — a technically elegant product that arrived into a market Microsoft was actively consolidating around Visual C++. JBuilder, Borland's Java IDE, addressed the emerging enterprise Java ecosystem and competed directly with IBM's VisualAge tools; it was capable, but Java developers were beginning to coalesce around open-source tooling, and the Eclipse project, released by IBM in 2001, would eventually make most commercial Java IDEs commercially marginal.

Kahn himself departed as CEO in 1995, and the leadership changes that followed reflected the difficulty of the company's position. In 1998 Borland renamed itself Inprise Corporation — a name intended to signal a pivot toward enterprise middleware and application servers — and the rebranding was received with widespread confusion and some derision from the developer community that had grown up on Turbo products. The Borland name returned in 2001, but the rebranding episode had cost the company credibility at precisely the moment it needed it.

Computer monitor displaying green text and graphs of system data on a dark screen

Green phosphor: the display most of Borland’s first customers were reading by.

Photo: Tima Miroshnichenko / Pexels

The decade of divestiture

The 2000s brought a strategy of managed retreat. Borland continued to develop and release Delphi — version 6 in 2001, version 7 in 2002, the latter widely regarded among the Delphi community as the last version to feel like the original product's direct descendant before the architecture began shifting. Later versions attempted to target the .NET framework, a transition that alienated a segment of the native-Windows constituency without fully capturing .NET developers who were already in the Microsoft ecosystem.

JetBrains had shipped IntelliJ IDEA in 2001 with automated refactoring as its central argument, and its quality of Java tooling steadily drew the enterprise Java market away from JBuilder. Borland discontinued JBuilder's commercial development in stages and eventually transferred it to an open-source model. Caliber, Together, StarTeam — the company assembled a portfolio of application lifecycle management tools, repositioning as an ALM vendor rather than a compiler company, a strategic logic that made sense on paper but moved Borland further from the technical identity it had built.

In 2006 Borland announced its intention to sell the IDE product lines — Delphi, C++ Builder, C#Builder, JBuilder — entirely, separating them from the ALM business. The sale took three years to complete. In 2008 the IDE assets passed to Embarcadero Technologies, which continued Delphi and C++ Builder development; the transaction closed formally in 2008. Borland itself, having shed the tools that had defined it, was acquired by Micro Focus International in 2009 as an ALM vendor.

The arc from 1983 to 2009 runs from $49.95 and a compiler that fit in RAM to a company that no longer owned a compiler. What Borland created — the integrated, affordable, fast-response development environment as a commercial product — had become the ambient assumption of the industry. The measure of that achievement was precisely that no one, by 2009, thought of it as remarkable.