Programming Windows: BASIC Inventors Take on Gates, Microsoft (Premium)

As noted in Programming Windows: Microsoft Basic (Premium), the BASIC programming language was created in 1964 by John G. Kemeny and Thomas E. Kurtz, two educators at Dartmouth College who wanted a programming language that was simple enough for non-engineers and scientists to use. Then, over a decade later, Bill Gates and Paul Allen developed a BASIC implementation for the MITS Altair and cofounded Microsoft to sell that product.

As genesis stories go, it’s a good one. And for those who care about BASIC, it’s pretty impressive that Microsoft continued supporting and advancing the language well into the 21st century. I’m particularly enamored of the work that Microsoft did with Visual Basic, which I feel respected the spirit of the language by making programming accessible to a new generation of developers and enthusiasts raised on GUIs.

So, I was fascinated—and a bit irritated—to discover that BASIC inventors Kemeny and Kurtz, when given a platform by Byte Magazine in 1990, used it to do little more than criticize Bill Gates, Microsoft, and the versions of BASIC that they had ridden to great success.

The irritation comes from the fact that Kemeny and Kurtz were at the time trying to sell their own commercial version of BASIC, called True BASIC, a product I had completely forgotten. It’s clear that part of their disdain for Microsoft’s BASICs is pure—driven by their love of their creation. But it’s likewise clear that these two disgruntled educators also used a very public soapbox to sell their own product too.

I’ll let you form your own opinion on whether that’s OK. But for the full context, we need to briefly discuss True BASIC first.

Kemeny and Kurtz had never copyrighted the BASIC language and they were taken aback by the different implementations that appeared throughout the late 1970s, which they called “street BASIC.” So, they sought to standardize the language, as would later be done with C and C++, in this case through the ANSI standards body. To do so, they added string variable support to their Dartmouth BASIC and branded the result as Minimal BASIC. Then, they added structured programming features, bringing it more in line with more sophisticated languages like C. This was named Standard BASIC (SBASIC).

The effort succeeded, as ANSI ratified both versions of BASIC. But by that time, Microsoft BASIC was already embedded in millions of different personal computers worldwide, making the standard pointless. Microsoft had, in a sense, usurped control of BASIC by creating a de facto standard. So, they withdrew their BASICs from standardization and decided to create a commercial product instead. This was called True BASIC.

True BASIC was released in 1985. It didn’t require line numbers or GOTO statements, and it supported advanced structured programming capabilities via subroutines with global and local variables. And in a very future-leaning move, True BASIC was designed such that it would work identically across platforms. True BASIC, Inc. created versions of its product for the Radio Shack TRS-80, Commodore Amiga, Atari ST, MS-DOS, Windows, and the Macintosh.

In tandem with the release of True BASIC, Kemeny and Kurtz also published a book, Back to BASIC: The History, Corruption, and Future of the Language, whose title provides a bit of insight into what they thought of Microsoft seizing control of their language. (I’ve not read the book.)

Regardless of the intent, True BASIC wasn’t successful, and it certainly didn’t knock Microsoft or its BASICs off their perch. (Even Borland, which saw great success with Turbo Pascal and Turbo C, couldn’t put a dent in Microsoft’s BASIC dominance. Its Turbo BASIC product, a repackaged version of BASIC/Z, was never particularly successful either.)

So, one might imagine that Kemeny and Kurtz were still a bit disgruntled when Byte Magazine came calling in 1990. As part of the same 15th-anniversary issue that gave us the insider account of the creation of the first IBM PC, Byte also published two editorials by the BASIC inventors. And they did not mince any words.

“I think that there was an unfortunate start for BASIC in microcomputers,” Kemeny said in the shorter of the two pieces (which are clearly transcriptions of conversations with the men). “Earlier in this decade, it was implemented on really tiny microcomputers. It was a miracle they could implement anything, and I admire what they achieved, [but] I think some really bad versions of BASIC hit microcomputers.”

Meaning Microsoft versions, in case that’s not obvious.

“I know when I first got an IBM PC, I loved the machine,” he continued. “But when I tried the BASIC that came with it, I was horrified. It was ugly and not at all [in] the spirit I thought BASIC should be. I think we’ve had an influence on that through the creation of True BASIC. I think our competitors have come out with much better versions [of their products]—I think under our influence. But I think that BASIC got enough bad publicity during that period that people have moved away from it.”

“I believe in two kinds of languages. When one really needs speed, one has to go to something like machine language, and I look at C as the modern version of machine language. So I can see writing some stuff in C, but [for] anything else, a language like BASIC—and I, of course, personally like BASIC. Good, structured BASIC, I think—and I know a lot of people at a lot of schools that feel the same way—is an excellent high-level language in which to [program]. And if in some parts of it you need extra speed, you switch to C.”

“In my experience, it is not the running time that is the overwhelming factor; it is the programming time that takes forever,” he concluded. “And writing in a good, structured language just saves you enorm0us amounts on programming, and on finding bugs later on.”

To be fair, Kemeny’s complaints were echoed by many of Microsoft’s competitors of the day: Gates and his firm were already infamous by 1990 for popularizing ideas that had occurred elsewhere. And many of the originators of those ideas were not necessarily happy with Microsoft’s less than elegant but successful attempts at selling those ideas to the unwashed masses. Still, this reeks of sour grapes to me. And, as noted, Microsoft deserves credit for sticking with the language that started it all, and for improving it over a long period of time.

To defend Kemeny, he isn’t just a crank. Elsewhere in the same issue of Byte, he has a quote that I thoroughly agree with, and always have.

“I still feel very strongly that learning how to program is an essential part of understanding computers,” he said. “I don’t mean someone who is just going to use it for word processing. But if you are going to do any serious work on computers, you really ought to program once in your life. I’ve even argued that if they are never going to write a program later in their life, the experience of writing a few programs, and particularly the experience of trying to debug programs, is a terribly important experience in understanding computers. Without it I don’t think people have a feeling on how much they should believe of what comes out of a computer.”

As for Kurtz, his Byte editorials is perhaps more measured. (Here, I’m only quoting the relevant parts, as his piece is longer.)

“I think BASIC is here to stay,” he wrote. “Anybody who has a personal computer probably has a bundled BASIC. This really includes practically all IBM PCs, which have some form of BASIC bundled with them. (Of course, that’s the line-number version. I’m amazed that people write programs with that BASIC.)”

OK, maybe not. That’s an obvious dig at Microsoft. (Microsoft QBASIC and QuickBASIC, which debuted in 1990, did finally remove the need for line numbers.)

“And I think the conversion from that version of BASIC to a more modern version will be very, very slow,” he continued. “[But] for people who do programming, it’s really asking too much of them to use Pascal or C. So I think that BASIC is here to stay.”

“What [Bill Gates] has done with his BASIC is to essentially turn it into Pascal,” he continued, turning explicitly to Microsoft. “You’ve got a lot of the attributes of Pascal. You’ve got data typing; you can declare new data types. And true, if you can do that, you can get rid of some of the uglies that are a part of Pascal, and that’s a sensible thing to do to a language that he wants as a systems programming language, a programming language for applications on the DOS environment. So he can put anything he wants in, so that it will be especially targeted for that environment. And that seems reasonable.”

So does his assessment of Microsoft’s evolution of BASIC, which is good. But it’s interesting how personal that got: Kurtz says “he,” clearly alluding to Bill Gates. As if Microsoft’s CEO of the time was toiling away, “corrupting” BASIC personally. (Perhaps the interviewer asked Kurtz specifically about Gates.)

“We would love to do [that] to our language, to add things like data typing and record structures and so on, but we feel that it would mess it up so much that we’d lose the flavor of what we have,” he concluded. “We’re a different environment—we’re a machine-independent language, so we cannot, [in] any appreciable way, put in features that are designed specifically to attack problems in the PC or the Mac. And it’s perfectly clear that those two environments are completely different in terms of system development. That doesn’t mean that you can’t use True BASIC in those environments, but they have to be applications that don’t require you [to] ‘bit twiddle’ and don’t require you to get at some specific hardware features.”

OK, that’s more short-sighted, perhaps, than horrible. In that limiting BASIC to support only features that will work everywhere is literally limiting and not an advantage. Which would probably be obvious to anyone that was a good businessman (Gates) as opposed to a college professor (Kemeny and Kurtz). But it’s not mean, at least. (And it is not coincidental that future cross-platform programming languages like Java would likewise suffer from their makers’ desire to work identically everywhere.)

But there is one more thing.

Amazingly, True BASIC still exists. (It is marketed as “the original BASIC,” which is a bit of a stretch.) And it’s still cross-platform—True BASIC 6 runs on “DOS, macOS, Windows [95 and newer], Unix, and Linux systems”—and still works the same across the supported platforms. It’s not clear what involvement that Thomas Kurtz—now 91 years old—still has in the company, however. (John Kemeny passed away in late 1992.)

I wonder what Kemeny and Kurtz thought of Visual Basic.