I hate April Fool’s Day. So nobody thinks this is an April Fool’s joke, I’ll just write more about what I wrote about yesterday, concentrating on media reactions to Paul Allen’s memoir. Then, tomorrow, I’ll revisit a very serious, important topic.WSJ: Would Gates have been successful without Paul Allen?
Yes, probably. But it wouldn’t have been Microsoft without him.
Allen says Gates wrote about 45% of Altair BASIC, the product that gave Microsoft its big break. Allen says he wrote 20%. But there’s one crucial detail in Allen’s account. Not counting the assembler macros and the 8080 emulator. Allen wrote those two crucial pieces. Did Gates know enough about the PDP-10 to write those two pieces? Perhaps. But without those two pieces, there’s no way Gates gets Altair BASIC finished in two months on his own. Allen may not have been able to spend as much time writing the 8080 code that went into Altair BASIC as Gates did, due to his job responsibilities at Honeywell. But being freed from having to write those two pieces was part of what allowed Gates to write nearly half of Altair BASIC.
I’ll see if I can put this in layman’s terms. Imagine if Gates and Allen had been writing a novel, and they had to use a particular word processor in order to do it. Gates wrote 45% of the novel, and Allen wrote 20% of the novel, but Allen wrote the word processor.
In more technical terms: Allen’s assembler macros allowed them to use the PDP-10’s assembler as a cross-assembler, causing it to output 8080 code for the Altair, rather than PDP-10 code. And Allen’s 8080 emulator allowed them to test the code to see if it would run, without having actual Altair hardware to run it on. It’s exactly the same way hobbyists develop for old game consoles on PCs today. They use a cross-assembler on a PC to generate code, then they test the code in an emulator, and once they’re satisfied with how the code runs, then they burn a cartridge.
Without Allen’s contribution, it would have been necessary to acquire an Altair. That was doable, but there was a waiting list at the time. It easily could have taken more than two months to get the hardware. They could have written code out on tablets while they waited for it to arrive, but they would have had no way to test it.
Ironically, without Paul Allen, about the only person Gates would have been able to turn to for 8080-compatible development tools would have been Gary Kildall. Gates’ and Kildall’s paths crossed a couple of times, first with Microsoft producing hardware to allow Kildall’s CP/M to run on the Apple II, and then in the competition to provide an operating system for the IBM PC, which Microsoft famously won.
Did Paul Allen deserve only 40% credit for Altair BASIC? Only 36%? It’s hard to say. But Gates told Ed Roberts that Microsoft was almost finished with the language before they’d even started. Ed Roberts wanted to see it in a month. Paul Allen flew out to demonstrate the language after about two months. That’s a remarkable achievement, but technically they were a month late. How receptive would Ed Roberts have been if they’d been two, three, or even more months later than he expected?
If none of this convinces you of the value of Paul Allen’s contributions, let me put it one more way.
I’m not a software developer by any stretch of the imagination. I took three programming classes in college, got As and Bs in all of them, but computer science passed me by in the early to mid 1990s. I can program a little in 8080 assembly language, a little more in 6502 assembly language, and a fair bit more in Microsoft BASIC, of all things. With Allen’s contributions, I would have been able to contribute a little to the project. Without them, I couldn’t have contributed anything of use.
I disagree with Dvorak’s analysis that Gates’ attempt to seize Paul Allen’s Microsoft shares is a trick to sell books. There are differences between Allen’s account and the account that Mark Stephens writing as Robert X. Cringely gave six years ago, but it’s close enough that it’s pretty clear they were writing about the same incident. This isn’t something that just came up out of the blue. Like I wrote yesterday, if anything, I think Allen toned it down. Allen’s quote, specifically, “This is unbelievable! It shows your true character, once and for all!” sounds contrived to me. It reads like George Lucas dialogue. Most people don’t talk like that, especially when angry. At the very least, he removed some four-letter words. And most writers know a made-up-after-the-fact quote when they see one, and that’s probably why Dvorak thinks this is made up. But enough other people have told the same story that I think something like that really happened.
Why the differences? These are the same people who can’t agree whether Microsoft competitor Gary Kildall refused to meet with IBM in 1980 because he was afraid of their lawyers’ NDAs, or because it was a nice day and he wanted to go fly in his airplane. To me it’s incredible that the accounts are as similar as they are.
Seattle Times: Microsoft employees are taken aback?
I’m not sure what’s in the published excerpts to be taken aback about. To me, this is like reading or watching Ringo Starr talk about the Beatles after reading stories in Rolling Stone. You’ve heard the stories before, but now you’re hearing an account from one of the participants. And it’s a little different, hearing it from one of the participants rather than from observers, but still recognizeable.
Is Allen bitter? I’m sure on some level he at least wonders about what might have been. But the more I think about it, I think he’s trying to answer the question of why he’s worth $14 billion while Gates is worth $56 billion. And the answer to that is twofold. One, Gates started with more. Two, Gates was willing to let a greater percentage of that wealth be tied to Microsoft. Is Allen bitter, or is he trying to explain why he was less willing to let his wealth be tied to Microsoft?
He didn’t want to tie his fate to Microsoft because he didn’t want Microsoft to be his life, unlike Gates for 20+ years of his life. And perhaps he didn’t want to tie up his wealth in Microsoft stock because he didn’t completely trust Gates. And would you blame him?
And why wouldn’t he expect someone would ask that question? After all, some of Allen’s investments have gone well, and some have not.
As for what the book will do to the relationship, who knows. They obviously weren’t best friends anymore but weren’t estranged either. Maybe it puts a strain on the relationship and maybe it doesn’t. But there’s no tougher test of a friendship than doing business together. I know that from personal experience. Some friendships can handle it and some just can’t.
The Seattle Times’ Brier Dudley writes, “Hopefully ‘Idea Man’ is more than a vendetta.” And I think it is. These stories aren’t new; the perspective is. And when it comes to talking about the early days of Microsoft, the three perspectives that really matter the most are Gates, Ballmer, and Allen. Allen has the least to lose by being completely candid.
If Gates ever writes his memoirs, I don’t think they’ll be candid. When asked how Microsoft got the PC DOS contract over Digital Research, all Gates ever said was, “Gary went flying.” Kildall didn’t live long enough to finish his memoirs, but they did serve as source material for a book chapter published in 2004. And Kildall’s side is much more complicated.