by Adam Fisher
Aleks Totić: We would have these fantasies. “Let’s have a company called “com.com.com” and live on a boat somewhere and it will be awesome!” And then it happens!
Jim Clark: I said, “Great. I will put three million in and I will hire the whole team.” I made myself the CEO and printed all of these stock agreements and offer letters, one with each guy’s name, and everyone signed them and we were off to the races.
John Giannandrea: Jim Clark was very clever, because he took all of the young people that knew anything about the web—literally flew around in his jet, picking up these people—saying, “Hey, I have a job for you in California!”
Aleks Totić: A week later we’re in California driving down 101. We saw Oracle, Sun, SGI, and I was like, “How come no one told me about this place before? This is awesome!” It was like the mecca.
John Giannandrea: And then he paired them with seasoned people from SGI. So the first twenty or twenty-two employees were a mixture of people right out of school who knew the leading-edge thing on what was going on with the web, and then also these seasoned engineers. The SGI DNA was there. And that was the magic that kind of worked.
Jim Clark: I did not have a financial plan, and there was no way I was going to take the time to write a financial plan. I was running on instinct about what the network effect could be. I thought, If we can get a couple million people using our product pretty quickly, there is going to be money to be made. The leap of faith that large numbers of people using your product is going to yield a profit does not seem to me like rocket science, but it did then.
Aleks Totić: We didn’t know how to make money. Our moneymaking vision was not as strong as our engineering vision.
Jamie Zawinski: “Give it away, and make it up in volume!” Well, no, originally, we had two software products. We had a web browser, and we had a web server.
Lou Montulli: Our original plan was to give the browser away for free, or for near free, and make all of our money on enterprise server software. We were going to give away the razors, and the razor blades were going to be our web server and our commerce server and all the other servers we were going to create.
Netscape’s give-away-the-razor-and-sell-the-blades strategy was almost immediately put into jeopardy. A rival team, Apache, was planning to make sure there would always be a free alternative to the server software that Netscape was developing—free razor blades for Netscape’s free razors. Apache wasn’t even a company. It was a group of part-time developers led by Wired’s webmaster, Brian Behlendorf, who took the hacker ethos—that information should be free—to heart. Behlendorf’s motives were as high-minded as they come, but his arrows were aimed squarely at Netscape’s business model.
Brian Behlendorf: There was an e-mail sent to this list of users saying, “Unfortunately we have bad news. The entire NCSA Mosaic development team has been hired by this new company called Netscape. The good news is that now the web will be commercial. Now the web will have official support. Now there will be a company that you can buy a product from, and that will sell you support, and if you have any problems you will be able to call a number. And you will get that all for the low, low price of a couple of thousand dollars per server.” It did not quite say that, but that was the implication.
Jim Clark: The instant response was, “You guys are crazy, and we hate you, because you are going to ruin the internet.”
Brian Behlendorf: We did not want the web to be owned by anybody, and felt that at least at the layer of web protocols, getting pages to people, we felt like the web server is like the printing press. All of us were running our own printing operations—building interesting websites, building interesting stuff. We just did not want to wake up one morning and find out that we had to start paying a tax to do the things that we had been doing for free before. It was very much an idealism kind of thing. It was inherited from where a lot of internet technology had come from, which is this idea that tech should be public and distributed. It is not necessarily the same point of view as saying, “All software should be free.”
Jim Clark: I said, “You do not get it. The only way the internet is going to be is if it is financed by businesses and operated by businesses, because the government cannot continue to put money into it ad nauseam. It has to be a commercial medium.” The gnashing of teeth! The wailing! And I got so much hate mail and nasty e-mails that you just would not believe it.
Brian Behlendorf: We were happy to see that these college students were getting real jobs, I guess, but at the same time a little sad, like, Well, okay. NCSA’s server software continues to work for us, and if we continue to fix bugs and add features to it, do we need Netscape? So I set up an e-mail list, and we started saying, “Should we give it a different name? It is obviously not the NCSA web server anymore. Maybe we should call it something else?” So we called ourselves Apache and figured out how to coordinate our efforts, we being a core team that was still ten or twelve people strong, even on day one, and grew to a couple hundred pretty quickly.
Meanwhile, back at Netscape…
Jim Clark: By June of ’94, everyone was showing up. There was already a team of eight to ten people, administrators. We were in full swing. Over the course of the summer, I began to worry about running out of money, because we were building up people. I only had maybe $15 million to my name, and I had already put in three. And so I did not have a whole lot more money I could afford to put in it. What we needed was to get the product done.
Jamie Zawinski: When I decided to take the job at Netscape, I thought I’d been working really hard. Turns out, I didn’t really know what “working hard” was yet. It was a lot of work.
Aleks Totić: It was great! Since we were all in the office all the time there was no calendar. You met several times a day and basically went over what’s going on and where are we going next. I think I was drinking twelve Cokes a day at that time. Or more. You lived on Coke and not too much else. Every once in a while when your body broke down completely, then you would take a day off and you would say, “I’m not coming in today,” and you would sleep all day. That would just go on and on and on.
John Giannandrea: General Magic was more intense, but the initial six months or year at Netscape were crazy. Jamie had a particularly hard job.
Jamie Zawinski: At the time, I was living in Berkeley and commuting to Mountain View, so I was only going home every other day. I’d either just stay awake for a day and a half, or while I was waiting for some code to compile, I’d take a nap in my chair.
John Giannandrea: He had an Aeron chair, an extrawide version. And he could sit cross-legged in his armchair, easily, and he would be programming with no shoes on. And he would have a pair of big headphones on playing thrash music of some kind, loud enough that you could hear it through the headphones. And when he was taking a break he would put a blanket over his head and sleep like that with the music going.
Lou Montulli: I have never had a more productive time in my entire life. A great deal of that was I could dedicate all of my energy to it, every waking moment, and not think about anything else because I had no distractions to pull me away at all. We were able to just get stuff done amazingly fast. I wrote about thirty thousand lines of code in three months. That is a ridiculous amount. I would not recommend it. It worked in that very specific circumstance.
Jim Clark: Toward the end of the summer, I wrapped up the financing, and we began to recruit some management to help. Then the University of Illinois began to rattle sabers, claiming we were violating their intellectual property, and implying that we were illicit and illegitimate. People at the University of Illinois had a lot of animosity toward us, because we were absconding with their crown jewels, and going to make money on them.
Lou Montulli: We were essentially doing a new implementation, a reimplementation, of something we already knew very well, Mosaic. Our goal was to create a brand-new web browser from the ground up. But we already knew what it should look like, and we basically knew how i
t should perform. We had all written the code once already.
Jim Clark: I had spent some energy trying to get everything smoothed out, and I wanted to pay them some kind of a fee just so they would be quiet. They would not take it. Well, they would, but they wanted it in the form of fifty cents a copy, and we were planning to give it away for free. I was not going to pay anyone fifty cents a copy when I was going to allow everyone to download it for free. That is a quick way to go out of business. Meanwhile, they gave Microsoft a paid-up copy for a couple of million dollars.
Aleks Totić: Microsoft gets the license; they have our old Mosaic code. We were worried. If we had our old code we would be really farther ahead than we are right now. But the old code sucks, so maybe not? The new Netscape code was much better…So it’s like our own code versus our own code.
Jamie Zawinski: We all believed very strongly that shipping fast was really, really important, because there were other people who were trying to do the same thing we were. So we had to be first, and we had to be best, that was our goal. “First” being the higher priority.
Lou Montulli: The last part of the release process is the most stressful, because it is constant back-and-forth with bugs and just trying to get this thing right, and it is the most frustrating time, because bugs sometimes take an indefinite, unknown period of time to work out, and you are like pulling your hair out and trying to find this thing, and then you have the a-ha! moment, and you do not know if the a-ha! moment will be ten minutes from now or a day and a half from now. It is really that kind of CSI moment: You have to find this thing within a couple hundred thousand lines of code, which is difficult.
Jamie Zawinski: It’s incredibly stressful. The days leading up to it were a lot of sleepless nights. The bug list is still long, but we have to decide if we are going to delay this release because of this bug, or not: If there’s a bug, and someone has a fix, we have to sit there and argue about whether it’s safe to put that fix in—is there a chance that this fix is going to destabilize other things and make it worse? So there’s a lot of waiting and panicking.
Lou Montulli: Jamie and I were very close, then I started to notice this more annoying side. Especially in meetings, he would just be really abrasive about things. He was famous for sending out flaming e-mails: “How dare you do this to me, you are wasting my time, and you are a complete asshole for doing this, and do not ever fucking do it again or else I am just going to get really upset!” So those sorts of messages, and almost to that extent of meanness. I know he is a nice person inside, but he just could not help himself on that sort of stuff. I had gotten several of these, and more often than not, he was directing his flames at me.
Jamie Zawinski: All of us who worked at Netscape in the early days were terrible to each other. We were all really abrasive people. We did most of our negotiation by screaming and insulting. Not necessarily the most healthy environment, but that’s how it was.
Lou Montulli: It was three a.m.; I was at the end of one of those double-day binges. Jamie had just done the same thing, and he had sent out a flaming e-mail and went to lunch or dinner or something, wherever you can get a hamburger at two a.m. I was just fed up with it. I virtually never do this, but I just sent back an e-mail that said, “Here is the problem, it was really your fault,” and then I was really not nice, and I said something like, “I really feel bad for all the people who have to work with you,” it was bad. I was just trying to give it back to him so he could understand how I felt, and this was right at the end of the release process so it had been boiling up.
John Giannandrea: Montulli and Jamie were needling each other. But Jamie was under so much stress, and is a very emotional person, and probably hadn’t had enough sleep.
Aleks Totić: Jamie—he was temperamental. He coded well. He was different from us. He was more flamboyant. We were just pretty much corn. Straight corn. He had the funny look, he had an image, he had a sense of style which was fairly foreign to us.
Jim Clark: He had half of his head shaved. Some stylistic statement on his part. I completely ignored it. It did not matter to me; I did not give a hoot. He was a great programmer, brilliant young guy. I do not think anyone bothered. People have just got to realize that a computer geek is respected on the basis of how much code he can write and the quality of code he can write. People do not give a crap what he looks like. If you generate good code quickly, no one cares. Jamie made everyone a lot of money.
Jamie Zawinski: We were there a lot, tempers ran high. Like I said, we were not friendly people to each other.
Lou Montulli: So it was boiling up, so I sent this e-mail and I went over to the cube next to me, to my friend Garrett Blyth, and I said, “Read it, read it,” and he reads it and he is giggling, and we hear Jamie walk in, and we are listening, and we hear him come in, log in to his computer, and then he starts typing really loudly, like really loud and vigorous, vigorous typing. Garrett and I know exactly what’s happening. Oh, this is going to be great! It’s going to be amazing flame coming back. I just cannot wait to read this. And we tried to take it in a good spirit, and right in the middle of the furious typing, we hear this sound. This very distinctive sound of a computer booting up. Or rebooting, in this particular case, for no particular reason at all. He is typing furiously and then his computer, for no reason at all, just reboots, which means everything that he has worked on just disappeared. So you hear the ding, you hear this slight pause, and then you hear, “What? What the fuck?!” and then he had this wall, this pyramid of Coke cans that was at least twenty layers deep, that was built up on this table next to his desk and it went up above the cube wall, and you hear this crash as all the Coke cans come off the desk in a sweep, and then this really loud sound to our left, and then Jamie storms off.
Jamie Zawinski: Lou was taunting me about something, and I lost it. I kicked a chair, and I walked out.
Lou Montulli: Garrett and I get up and we turn the corner, and there is Eric Bina with a chair like literally jutting out of the wall right next to his head. Apparently what happened is, Jamie got up, picked up his chair, and threw it over his head out of his cube. Eric had come around, because he heard the Coke cans, and the chair went right by his head and embedded itself into the wall. It was quite a scene.
John Giannandrea: He threw a chair and we sent him home.
Jim Barksdale: It was a big deal. We even had T-shirts printed up that showed the chair in the wall!
Jamie Zawinski: We made a T-shirt that had a coat of arms, and there was a chair in the middle of it. Yeah, because we were jerks and we taunted each other that way. That’s it. That’s the whole fucking story.
Lou Montulli: After that incident, he essentially went into his cube behind this thing—this large camo thing which you use to cover a tank or a plane—that was hung up over his cube, and he never talked to any of us ever again. It was kind of the end of the social world, at least between me and Jamie and the core team. It was a highly charged event; I do not fault him for any of it. I certainly provoked it as much as he did. It was right at the end of our release process—the beta release, the very first beta release. We had been working four or five or six months, seven days a week, 120 hours a week, and all of our core motivation was to get it out into people’s hands because if nobody is using it, it is essentially mental masturbation, right? The end result has to be this thing that people can use.
The first Netscape browser was released into the wild—as a “beta” or experimental version—in October 1994.
Lou Montulli: So we are very very excited, and we all got in a conference room to celebrate the release. We were going to watch as people downloaded it live, and we kind of figured out how to do that.
Jamie Zawinski: There’s a log file showing each download, and one of the guys rigged up a little script to parse that log file as it was scrolling by.
John Giannandrea: I don’t know why, I just did it, and people thought it was kind of cool.
Aleks Totić: We had a different sound e
ffect for different downloads—each different browser got its own sound bite. There was a frog, a cannon, and shattered glass. We’re in a darkened room, kind of watching the monitor log, and there’s nothing on the log. Then the first download came from Australia. It took him a long while. He got half of it then he had to restart. And then it finally went: Boom! The first download. Cheers, champagne, beer, and all of it. A few minutes later the next one comes, and then a few minutes later the next one and it gets faster and faster.
John Giannandrea: It started off as a trickle, and then it became a flood, and eventually it’s just deafening.
Aleks Totić: Six hours later we’re still there and it’s like boom, crash, boom, crash! It’s an avalanche, and it’s never slowed down since.
Lou Montulli: I don’t want to toot my own horn, but it was really good software. We were very proud of it. In the modern experience, going from any other browser to Netscape would be the equivalent of going from your flip phone to an iPhone kind of experience, like, Wow, this finally works the way it is supposed to work! We were ten times faster, we had a lot more features, it was a lot more polished. Images became a lot more practical on a web page. Before Netscape, you would wait minutes before you saw anything. We changed that completely. With Netscape images would appear ten times faster, because the content would all stream in dynamically.
Steven Johnson: So what had just been text-based pages suddenly had images. This was the point which the web stopped just feeling like an interesting new format for data, and really started to look like media. You could start to imagine that this was something that could become a magazine, that you could put advertising on, or maybe the kind of intimate media of your own personal pictures. I remember looking at Netscape and thinking, Oh, I can learn this and publish my own magazine—with basically no distribution costs! That was just mind-blowing.