iWoz (23 page)

It was the first and most astounding success for Apple the company. I will never forget that moment.

• o •

Well, I decided I should rim the whole thing by HP one more time. I spoke to Pete again. He told me to run it by legal.

The legal department ran it by every single division of HP. That process took about two weeks.

But HP still wasn't interested, and I received a note from HP's legal department saying they claimed no right to my design.

• o •

It turned out that a guy named Paul Terrell was starting a new computer store, called the Byte Shop in Mountain View.

As I said, Terrell had seen me demonstrating my computer at Homebrew, and he'd told Steve to "keep in touch," and Steve followed up with him the next day. Steve showed up barefoot at his office the next day, saying, "Hi. I'm keeping in touch."

What Steve didn't know was that Paul was looking for a product just like ours. Terrell wanted to sell a complete computer to his customers, fully assembled. And that had never been done before. Before us, Paul had been buying Altairs or kits like that, and had technicians soldering them together in the back. Every time he got one built, he could sell it. But he thought he had a lot of interest, a lot more potential customers. Steve told him about

the Apple I I'd designed and Paul realized it was a completely built board, so it was a great product for him.

So suddenly, with Terrell's order, I could see that someone else was interested in the Apple I. That was so unexpected and exciting—and so easy. I mean, we already had a little company that was set up to mass-produce our boards down in Santa Clara. Now, all we needed to do was supply the additional parts and they would solder them on.

But how would we get the parts? That would cost money we didn't have. Allen Baum and his dad, Elmer, loaned us $1,200 to buy some of the parts. But we did end up finding a chip distributor (Cramer Electronics) and got the parts on a thirty-day credit. The chip distributor had to call Paul Terrell to see if he was really going to pay us.

The deal Steve worked out with Paul Terrell was he would pay us cash on delivery for the computers. So Paul Terrell was really financing this whole project, it turned out. When he paid us, we were able to pay for the chips.

The distributor gave us the parts, and then they went into a sealed closet at the Santa Clara company that was manufacturing the boards. On the day they were ready for them, the parts came out of the closet, were accounted for and soldered on, and then we had thirty days to pay for them.

Our first batch of boards was finished in January 1976. There were kits like the Altair out there, but nothing like what we were doing. I remember how, waiting for them, I was just the happiest person in the world. I was so happy around this time. I never truly thought we were going to make money with Apple. That was never in my mind. The only thing on my mind was, Wow, now that I've discovered what a microprocessor can do, there are so many places I can take it. I knew that for the rest of my life, I would have a computing tool for myself.

The potential with the Apple I was blowing my mind. I mean,

I'm around video games, and suddenly I realize that my little computer is going to be able to play games. I imagined word- processing software replacing typewriters someday. I was a fast typist, and I could see we were nowhere near where we needed to be for a computer to replace a typewriter yet, though I could imagine it. I imagined how a computer could help me with all my design work at HP. It just blew me away. Every single thing I thought about on the computer was going to be valuable. I could see it so clearly. And that was all I could think about.

After the boards were finished, we rounded up Steve's friend Dan Kottke and Steve's sister, Patty, to plug chips into sockets for $1 a board. Steve would bring us maybe ten or twenty assembled boards at a time from the manufacturer. And there we would sit on a lab bench in the garage of Steve's parents' house at 11161 Crist Avenue. Then I would plug each assembled board into the TV and keyboard we had there and test it to see if it worked.

If it did, I put it in a box. If it didn't, I'd figure what pin hadn't gotten into the socket right or what circuit was shorted. I'd fix the bad ones and put them in the box. After a dozen or two were in the box, Steve would drive them down to Paul Terrell's store and get paid in cash.

These weren't finished computers as you would think of them today. Paul Terrell ended up having to supply monitors, transformers, keyboards, and even the cases to put the computers in. I'm not sure that's what he expected. I think he thought, based on what Steve Jobs told him, that he was getting a fully built computer.

Back then, we didn't have the volume to do plastic. So Paul would put them into wooden cases—often a Polynesian wood called koa—which was a style thing for us.

We had to come up with a retail price for our literature. After all, we weren't going to sell them just to Paul.

We decided to price them at $666.66 each—a price I came up with because I liked repeating digits. (That was $500, plus a 30 percent markup.)

And you know what? Neither of us even knew the number's satanic connections until Steve started getting letters about it. I mean, what? The number of the Beast. Truly, I had no idea. I hadn't seen the movie
The Exorcist
. And the Apple I was no beast to me.

• o •

By now, writing the BASIC interpreter was turning out to be the longest, most complicated single project I'd ever do for Apple.

Man, I sniveled at BASIC back then. Compared to FORTRAN, it was a weak, lightweight language. I thought no one would ever use it, for example, to create the kind of sophisticated programs engineers and scientists use. I could just see where tilings were going. That book I told you about,
101 Basic Computer Games,
meant you could just type in the programs and have these games.

I'd been writing a BASIC interpreter to run on the Apple I, which was based around the MOS 6502 processor. I figured if I wrote this language really fast—if I worked on it day and night and turned my ideas into something that worked within a couple of months—well, then I would get almost famous. People would say that Steve Wozniak wrote the first BASIC for the 6502, just like they knew Bill Gates for writing the BASIC for Altair. I would be the source, and that was kind of exciting.

I had never taken a class on writing a computer language. In my early college years, Allen Baum would Xerox textbooks at MIT, where he went to school, and send those pages to me. I learned a little that way.

So I understood that computer languages had a grammatical syntax, just like any language, and I knew how they were organized.

I didn't know that the BASIC interpreters that existed for different computers, like DEC's and HP's, were different in any way.

I assumed they were all the same, and I assumed Bill Gates's was the same as those.

So, back at work, I grabbed some HP BASIC manuals and studied them. I started writing on paper a syntax table. This is what describes the grammar of the computer language. It defines what commands a programmer can enter.

For instance, if English had a syntax table, it would explain that personal pronouns like "he" and "she" are nouns and usually subjects in a sentence such as, "He threw the ball." It would list all the possible verbs, of which "threw" would be one. And it would tell you what all possible "objects" would be, such as "ball." In English, there are millions of possibilities for subjects, verbs, and objects, but in a language like BASIC you can limit them to a certain number of items.

Then there are the rules you need. Say you wrote out the equation 5 + 3x7. When you write that out with no parentheses, a mathematician would know that you execute multiplication and division first, addition and subtraction second. So really that equation would be 5 + 21. So that rule, about what terms to execute first, is an example of something that has to be defined in the syntax table.

I had no idea what other people did in their computer languages, but I felt it was obvious that you needed a noun stack to hold things like numbers, a verb stack (which would include actions like multiplication or addition), and a set of priorities for every single verb that was typed in.

It took me about four months to come up with the core of my BASIC interpreter. I ended up leaving out the ability to type in decimal numbers (called "floating point arithmetic"), and instead handled everything with integers—that is, whole numbers. That saved me about a month of work, I figured. I decided that for games and computer simulations—the two main things I was writing the BASIC for—I would just get by with integers.

Many of the key programs in my life, including those back in Colorado, used only integers. So I designed my BASIC to only work with numbers from -32,768 to +32,767.

I wrote the whole program on paper—with machine instructions on the left and the equivalent in hexadecimal (equivalent to Is and Os) on the right. I had to do this by hand because I couldn't afford to use an assembler program, which is the typical way you'd do this. This is the same way I had to write the little monitor program.

So I figured, Hey, I'm able to write the program with the code by myself, by hand. Who needs a computer to do this for you?

By the way, I still have the notebook I wrote my BASIC interpreter in. I'm not sure, but I bet it could be worth a lot to a museum now!

Anyway, the end result of all this was, when my 6502 BASIC was in the computer, I could type in little programs with the keyboard. Like I could have the computer ask you, "What's your name?" And if you typed it in, it would fly your name all around the screen. This sounds so simple now, but back then, nobody had ever seen a small computer where you could actually type in programs with a regular keyboard and have it execute.

Even with machines like the Altair, it was very expensive to add a plug-in card and a big cable that would connect to an enormous ugly teletype that had the keyboard to type on.

I showed my computer running BASIC at Homebrew after the main meetings a few times, and people were just blown away by it.

But there was one problem. The Apple I had no permanent storage—no hard disk drive or floppy or CD drive like you see today. This was way, way before that. So every time I wanted the BASIC to run, I had to turn on the computer and literally type my whole program in from my notebook. This was a 4,000-byte program—it took almost forty minutes for me to type it all in every time. And when I turned off the computer, because there was no

The Basics on BASIC

The BASIC computer language, the one I told you I sniveled at, was designed from the start to be an easy-to-program language for writing computer programs. Created in 1963 by Dartmouth College professors John Kemeny and Thomas Kurtz, the BASIC acronym stands for Beginner's All-purpose Symbolic Instruction Code. There's no question that BASIC is an easier-to-learn language than languages such as Pascal or C. It's also smaller and slower. But it worked just perfectly with my first Apple computers.

permanent storage and only RAM, the whole program would be gone. I ended up having to either leave my computer on all the time—which meant I couldn't transport it very well—or come up with a solution.

That's how 1 developed the cassette tape interface for the Apple I. Aside from changing the type of RAM from static to dynamic, this was the only change to my original design from the very first days of Homebrew in the spring of 1975. I designed a circuit so that a regular cassette tape could hold the BASIC, and when I turned the computer on, it would automatically load my BASIC into memory so the computer was usable.

Once the BASIC was done and easily loadable from a cassette tape, I discovered something terrible. I had miscalculated. I had thought that all versions of BASIC were more or less the same, and that all the 101 games in BASIC that I had in that book would automatically run if you typed them in. That turned out not to be the case. It turned out that the type of BASIC I'd written—as well as the HP BASIC I'd originally studied—was totally different from Bill Gates's Microsoft BASIC, which was based on the DEC BASIC at the time. Bummer!

So anyone who wanted to put those games on the Apple I was going to have to make changes to the games to do it.

I did manage to get some games working on the Apple I, though. There was a popular game, in BASIC, called Star Trek. Like the show. I adapted it to my BASIC and it ran just fine.

• o •

After we started selling the boards to Paul Terrell—working day and night to get them to him on time—we had profits like I never imagined. Suddenly our little business was making more than I was making at HP That wasn't very much, admittedly. But still, it was a lot. We were building the boxes for $220 and selling them wholesale to Paul Terrell for $500.

And, of course, we didn't need a ton of money to operate. I had a day job, so I looked at it as, Hey, cool. Extra money for pizza! As for Steve, he was living at home. I was twenty-five and he was only twenty-one at the time, so what expenses could we have, really? Apple didn't have to make that much to sustain itself and be ongoing. We weren't paying ourselves salaries or paying rent, after all. We didn't have any patents to pay for. Or lawyers. It was a small-time business, and we weren't worried that much about anything.