Microsoft Altair BASIC legend talks about Linux, CPRM and that very frightening photo
A very rare interview with Monte Davidoff
Twenty six years ago the microprocessor revolution found a software catalyst - a tiny BASIC interpreter that ran in 4K of memory. You've probably heard of two of its three authors - Paul Allen and Bill Gates, who'd incorporated the company 'Micro-Soft' in Albuquerque the same year. The third man, Monte Davidoff, isn't nearly as famous. You'll search in vain for an interview on the web with Monte. So we figured we'd set that right.
Davidoff contributed the maths routines to the interpreter, and we finally tracked him down in Cupertino where he's enjoying the recent freedom of being an independent software consultant, as Alluivial Software.
With so many questions, we figured we'd start off with the most important. Was he in the famous Microsoft team photo from 1978 - the scary one where almost everyone has terrifying arrangements of facial hair except for the young CEO - and could he please identify himself? As it happens, he wasn't there, but he does shed light on its origins:-
"That was taken after my second summer, working on the second BASIC for Microsoft. The story there is that Bob Greenberg (center right) had won a prize from a photo lab, which was a free photo shoot." So the first corporate publicity shot came about completely by chance.
And does the scurrilous rumor that Allen and himself did all the work, while Bill played poker hold true?
Not at all, he says. "We were both working pretty hard. The maths routines are a part of the interpreter, not all of it, and Bill and Paul wrote the rest."
The software was lauded for its efficiency and small foot print. But execution time wasn't uppermost in the authors minds, says Monte:-
"We weren't too concerned about efficiency. Even the 8088 then was a pretty fast processor. And we were running it over an ASR 33 Teletype running at 110 baud!"
The main constraint, back then, was memory.
"It had to run in 4k. In fact the 8k version had algorithms that were more efficient but that took up more space. By the time the 4k BASIC was done, the 8k version was out." Incredibly, the three of them produced the interpreter without seeing the MITS Altair itself - the coding and debugging was done entirely on a simulator.
Along with other pioneers of the time, component shortages weren't as challenging as what they could do with the chip, says Davidoff.
"The memory producers were churning out memory at a constant rate - what was new was the microprocessor itself. The 8080 wasn't the first - Intel had the 8008 and the 404 before that - but the 8080 was the real breakthrough."
Davidoff was a Harvard student at the time, writing the BASIC interpreters over two summers in 1975 and 1977, and returning to study Applied Maths and Science. From there he worked on the Multics for Honeywell in Cambridge, Mass. and then at Tandem, HP's Research Labs, and Stratus. Which he stayed with for twelve years, through acquisitions by Ascend and Lucent.
Ah, Tandem. So did he work with Jim Gray, who's now head of Microsoft's Bay Area research lab and helped create the first relational database (System*R) while at IBM, and pioneered clustering while at Tandem, and later at DEC?
"He reviewed some things I worked on," says Monte, "but we were working on different product lines." That labor - file system work for a successor to the Tandem T16 - never saw the light of day.
Although he helped put the fledgling Micro-Soft on the map, Davidoff has subsequently worked with Unix for most of his career. Microsoft actually bought into Unix very early on in 1979, but its own AT&T derivative Xenix found few buyers, and it eventually spun the work out to the Xenix authors SCO. And these days, Davidoff runs Linux (Red Hat 6.1) at home.
"I'm really excited about Linux," he says. "Having used Unix all these years and put out professional Unix products, they've done a really good job."
His other passion, he tells us, is Python.
"Hats off to them. It's an extremely well designed language. It's object orientated from the get-go. They've really succeeded there," he says, and commends it as the ideal teaching language. That used to be BASIC, of course.
RMS and GNU
With Microsoft investing so much in fighting software libre recently, what we wondered, did he make of the schism between the proprietary and free software models?
Well, the different cultures were there from the start, he says. "Richard Stallman came from an academic background, a freedom environment. Bill had already had paid jobs doing software work for other people. What spiked controversy was that Bill wanted to bring payment into it.
But from his own experience the value of GNU's work has been consistently undervalued, he says.
"I have to say that from being in proprietary software so along, Stallman's contribution is enormous. Most managers don't fully recognize how much value they have got from GNU software. Particularly gcc [the GNU C compiler] - every embedded software uses compilers based around gcc. They've saved a lot of money, and don't even know it."
Monte was wary about drawing generalizations about the process, but pointed out that where one or a small number of people was in charge - citing Linus and the kernel, or Guido von Rossum and Python, or Apache - the end result was demonstrably better.
How copy protection could kill the PC
Davidoff was witness to the birth of a microprocessor revolution that provided almost limitless freedoms for users. And now he's here for what could be its funeral. The freedoms are at stake now, under threat from interests outside the technology industry for the first time. Hale Landis, veteran of hard disk standards, suggested as much in a posting we cited here, recently. And what he was referring to, CPRM, is just the most notorious, or the most emblematic of a number of schemes that make the open personal computer into an limited and tightly-controlled playback device. Controlled, effectively, by the entertainment industry.
Was the threat real, we wondered?
Absolutely, says Davidoff. "It's changing the way we're doing our own business."
And it's scarier because most people don't know about it:-
"I don't think people are aware of it, in spite of what you and others are writing about. It hasn't made it into the public consciousness," he says. "I didn't hear about the DMCA until after it had been passed."
"The entertainment industry has been made happy in the past. It's worked out some equitable settlement, such as radio rights, or a levy on blank cassettes. But it's got to be done without implementing a technology stranglehold, so people's data - that which is theirs - belongs to them."
Davidoff only has occasional contact with the Altair BASIC co-authors. He last ran into Gates "seven or eight years ago" but has been in touch with Allen. Ending up with a cornball question, we asked if he imagined 'Micro-Soft' would become as successful as it has.
"No. Paul and Bill were into the business, that's what they talked about, but no one could predict that they'd be this big."
Now, you might wonder, where can we see this historic code? Well, that dear readers, is an extraordinary tale in itself, which we can reveal here. ®
Sponsored: Benefits from the lessons learned in HPC