Tab Atkins: Hello again, everybody. Hope you had a great time in the sessions that we just concluded. Coming up next is a dive into the history of the web with Brian Kardell and Chris Wilson. We've all been working on the web for a long time, but not all of us have been around for that long. So this should hopefully be a really informative and fun dive into just what our past was. Thank you.
Brian Kardell: All right. Hi, this is the third blink on web history talk. It's actually the fourth web history talk in a way though, because the first one was a long form, lecture style talk that was given at Chrome University by Chris Wilson. It's a great talk, we'll link to it and we'll try not to cover too much stuff that's in there, but I am joined here today with Chris Wilson. Hi.
Chris Wilson: Hi. Thanks Brian.
Brian Kardell: Chris, you've been involved with the web since the very, very beginning. You worked on the first popular web browser, Mosaic.
Chris Wilson: Pretty much. Yeah. I started working on the web in 1993. I was working at NCSA doing other projects. I was the lead for the PC software and we started working on NCSA Mosaic not that long after the X windows version got started.
Brian Kardell: I think probably a lot of people who are listening don't really know what NCSA is. So can you give a maybe a 10 to 30 second answer to that?
Chris Wilson: Yeah. So NCSA stands for the National Center for Super computing Applications, and basically the university of Illinois through NCSA was one of the national science foundation, I forget what they were, funded sites basically. And they had several supercomputers and the group that I was in there, we were the software tools group. And our charter was basically to build tools that regular researchers, academic people and everything they could use to access the power of the super computer from the desktop that they probably had, like PCs, Macs, X windows work stations. So the first thing I was hired to work on as an undergrad was telnet for DOS, because literally DOS didn't have a network package. It didn't have telnet, it didn't have the terminal program built in. You had to go buy it or get free software to do it, and that's what I worked on, was that free software.
Brian Kardell: Yeah. I have difficulty imagining what it's like living in another timeline where I'm born 10 years later or something like that, because it must be really hard to relate to but there was not really Windows.
Chris Wilson: Not really. I mean, at that point we had Windows 3.11 was the Windows for Work Groups was the current version. And I mean, the key here was you still, even in Windows until Windows 95, you did not have a TCP IP package built into Windows. So the next project that I worked on, or one of the projects that I worked on before Mosaic was telnet forWwindows and we had to port the entire network package literally the device drivers for hardware network cards in order to get it up and running. We didn't have a sockets library to build on, we wrote the sockets library.
Brian Kardell: A very tiny percentage of American homes, and I think America was way up on the curve compared to the rest of the world, had computers at all. And a shocking percentage of those, it's not shocking, but you wouldn't think about it, even though Windows was out, this is a very strong likelihood that even if you had a computer, it didn't have Windows.
Chris Wilson: Or you didn't have a current version or you didn't use it much. I mean, certainly you didn't game from within Windows, you probably didn't game on your desktop computer at all. But Windows was something that people were just starting to use.
Brian Kardell: And how many of those people then also had a modem? Well, that was an optional thing.
Chris Wilson: Yes. It was pretty optional-
Brian Kardell: And then how many of those people could figure out how to use a web browser? It was a pretty niche thing when it started out, right? You had to fall into a relatively small niche to be a early person on the web. And you were at NCSA just right in the hub where you could-
Chris Wilson: Oh, I was absolutely in the right place at the right time. I mean, it was great to be there in the middle of it all and see. I mean, I grew into this whole network and internet and web not at the very beginning of the internet, but at the very beginning inflection point where you started getting content and other people connected. And that was when it really started getting interesting.
Brian Kardell: And so what's really interesting is at that time, the network protocols were, there were a lot of them, they were competing and. It wasn't even unique, some of the things that the web did. So in 1991 Sir Tim Berners-Lee released his, he wasn't a Sir yet, released his browser on NeXTSTEP, which was a really cool operating system that most people don't know. But his idea was in a lot of ways very simple and lesser than many other things that already existed. It had a lot of challenges, right?
Chris Wilson: Oh yeah. I mean, I think he made some interesting choices in the underlying syntax and things like that, but he really, he made the source as easy to understand as possible as human readable and writeable as possible and realized very early on that hyperlinking was really the biggest, most important piece to the whole puzzle of getting people hooked in. And he made the software so easy to build that you could easily run a web server. You could easily run a web browser and he didn't try to envision, this is the piece of software everyone will use to access this thing. He set it up so other people could go build that software too.
Brian Kardell: Could you tell me how closely could you have predicted that the web would be the thing that it is today, back then? Did you imagine?
Chris Wilson: I don't think any of us really imagined what the web would be today. This is so many layers removed from that. I think we all had a strong belief that it was a super useful tool. Really, it was going to be an absolute game changer. Not even a game changer, a completely different game from then on in computing, because it opened up all of this accessibility of content, of information. And I went into this in my talk a little bit, but it's impossible to really understand how different the world was before that.
Brian Kardell: I mean, even if you could see the potential for it, what's interesting about it is that the web was very much an underdog in this. It was very, like I say, niche and academic, and it didn't really have the big business bets behind it. Gofer was in the library systems at this point. Pretty much every library that I went to was on that, but you also had high time, there was a lot of buzz around High Time. High Time was going to be the real thing that had Ted Nelson involved in it. And I think Gofer, they were SGML ISO standards at that point. And it just looked like a lot of things about the web seemed academic and a little bit almost like a cute toy because of the simplicity.
Chris Wilson: I think the interesting bit is there was also this massive watershed that happened as bandwidth became more and more available and easier to come by and memory and computing power became easier and cheaper and more common to come by. I mean, I ran across something the other day that was an old ad for a computer, the Raytheon 704 that was released the year I was born. And I mean, keep in mind, this is not a personal computer, this is a corporate level $70,000 in today's dollar computer and ran at one megahertz and had eight kilobytes of memory in it, total.
Brian Kardell: Nice.
Chris Wilson: And I mean, even if you think about when I was in college and when I was working at NCSA, the power that you could get out of the supercomputer at that point, I have that in my cell phone, I don't even have the latest and greatest cell phone, but it's just the amount of memory. And we all watch streaming video today. And I remember a point in my career programming that it was going around basically like, oh, well, we can never get that many bits displayed on the video bus that fast. We can't animate full screen video in full color because it's just too much bandwidth. And it's like, well, wait a while and those things will happen. And I think now we're at this inflection point of, well, what can we do with more power or should we do? And hopefully we'll start answering that question more.
Brian Kardell: So maybe some of the web's initial weakness was actually a strength because it allowed a lot of people to get and help envision its future. So that I feel is what happened with Mosaic, right?
Chris Wilson: Yeah, absolutely. I mean, to be clear, the team working on Mosaic, we weren't the most seasoned software engineering superstars of the time. We're largely college students, a few grad students, most of us had either just graduated or were still undergrads and we went and built the code that kicked this whole thing off. I mean, I actually, I looked back at that part of the code and I didn't have a manual. I couldn't go look up on the internet, how to write code to do a rendering layout engine or anything like that. So I made it up. And I actually, when I eventually went to Microsoft and started using the Spyglass code, which was based on the NCSA Mosaic code, I could still see the design that I had that I'd originally put into NCSA Mosaic for Windows, because there was a very specific bug in the way that it worked. And I knew it was a shortcoming, but I was like, 'Ah, I'll fix that later.' And never did. And you see how those effects happen over time.
Brian Kardell: What's interesting is Mosaic was available on three operating systems, right? So they were not the same team though?
Chris Wilson: They were not the same team and most definitely not the same source code. I think we all started from Tim's libw3 to do the parsing of HDP requests and the basic parsing of HTML, but then we just hacked the heck out of them. I mean, at the time we were in 16 bit Windows and 16 bit Mac operating system, and they were so different in how you had to do things like manage memory that it would've and difficult, not impossible, but difficult to have shared code. And there wasn't enough impetus to actually do that because it was so visual. And not visual in a, okay, now you have to display the pixels way, but in a, okay, you need to scroll Windows. How does that work on Mac versus Windows or the like, and it was pretty, pretty wild.
Brian Kardell: So Mosaic made the web popular enough that a lot more people could try it. It got written about a lot. And then it split up. A lot of the people left to go other places. You went to SPRY, right?
Chris Wilson: Yeah. So I was actually looking to move on. I was getting married and my wife wanted to go to grad school and we decided we wanted to move out to with Seattle, and I'd connected with this company that was licensing the Mosaic source code, SPRY. And so I left NCSA, gave my notice. And I think this lit a fire under Mark and Grayson who had already left NCSA at the time and was talking with Jim Clark about starting this company that became Netscape. And they basically came back and interviewed most of the rest of the Mosaic development team and scooped them all up all at once. And then I only stayed at SPRY for a little over a year, I think, before I moved to Microsoft.
Brian Kardell: I feel like the generation of things like... So for anybody who doesn't know, SPRY made a thing that was internet in a box.
Chris Wilson: Oh yes.
Brian Kardell: Oh, you have it?
Chris Wilson: Mosaic in a box. right here.
Brian Kardell: Okay. Yeah.
Chris Wilson: There's a bigger product that was internet in a box and had all of the telnet and email and everything. We had all the clients. Because it wasn't clear at the time that mosaic, that the web was the winner. So literally, we had an email client, an FTP client, a Gofer client. And when I was working on the team, there was another one of my peers working on the Gofer client. We were investing in it just as much as the web browser.
Brian Kardell: An interesting anecdote to this is, this is about the time that I personally came to the web. I learned about the web in a bookstore because well, if there wasn't a web to learn about the web on. And I immediately bought a bunch of books. And one of them was an early book written by some fortune 500 consultant, whatever. And in it, he talks about how Netscape, this brand new startup, they just coined the market and Bill Gates lost. It's over. That's pretty interesting because then you went to work for Microsoft.
Chris Wilson: Yes.
Brian Kardell: And you worked on, was it ie3 was the time you came in or ie2?
Chris Wilson: So I actually joined the ie team, I actually went to a Microsoft and started on a different project altogether. When I first went there, I was working on, it was basically a distant forerunner of Bing search. I wrote a web crawler for them, basically. It was the only thing that I did there, but it was fascinating because I had to crawl office content, which was neat. But then the Ii team found out that somebody with a lot of browser experience was there and they came over and poached me, which was an interesting because at that point I was like, 'Yeah, I want to get back into the browser stuff. I miss that.' And I made some not so good friends there by jumping ship. But I joined shortly before we shipped ie2.0, I did check in and ie2.0 and then really got going in ie3.
Brian Kardell: I feel like since we're speaking at a Google event, it feels a little weird to say this, but this is very hard to appreciate how different the web was before Google. Search engines were much less,. And finding anything on the web was very, very difficult and frustrating.
Chris Wilson: A good example is actually the search engine that I was working on, we basically, we got the database, the index, we licensed it from Lycos, which was another early web search engine and then Microsoft put the content on their own database backend because they thought their database was just that much better and people would use it more. And the interesting bit was we got a new copy of the database once a week.
Brian Kardell: Wow. Do you remember searching? You'd have the search engines like Alta Vista, we imagine you would have 47 input fields where you would be like, 'I know who the author is' and was very much like trying to apply library science somehow to-
Chris Wilson: Pretty close. Yeah.
Brian Kardell: But it was very interesting, we good should do a whole podcast about that at some point or something. But so anyway, you went to Microsoft and many of your coworkers went to Netscape. And you made competing engines.
Chris Wilson: Yes. I think part of that was, I had actually already made the commitment to move to Seattle at the time and I liked living in Seattle once I was there. And when I left SPRY, Netscape was doing some interesting things, but there was definitely a strong attitude to them as well. And not one that I'd been super fond of when I worked with a lot of the same people, not all of them, this isn't a generic statement on people, but I wanted to try something new and that's how I ended up at Microsoft. It was more I was looking for a job and wanted to move on. And I think that it ended up being an interesting dichotomy because were talking about this earlier. We were competing like our companies definitely were strongly competing, but the people who were working on this stuff, we were collaborating too. We were trying to build this platform together to some degree. And there was more positioning, I think, than there is today about we would come up with some new feature and we would build it and then we would ship it and then we would propose it in a working group or something. And that's not the kind of thing that's okay to do today certainly. I think we've gotten away from that, which is good.
Brian Kardell: Yeah. That's an interesting thing that I actually wanted to ask you about was I think the browser wars was a thing and for developers, it was simultaneously really exciting and really, really frustrating. So you could do amazing things as long as you only wanted to serve as one browser.
Chris Wilson: Absolutely.
Brian Kardell: And actually either browser, you could build amazing things with really, you just would have to do it very, very differently. And you could do a little bit more over here than you could do over there on one thing, but not on the other thing. And it was unfun in a lot of ways, but ie6 was at the time it was released undoubtedly the best browser.
Chris Wilson: Yeah.
Brian Kardell: It was hands down better. I'm trying to think, early browsers before, it might have been before ie6, there were no DevTools. I think Venkman on Mozilla was the first one I ever used.
Chris Wilson: Yeah. I think that's definitely in the lead for real developer tools.
Brian Kardell: Yeah. So browsers have come a long way. They've gotten much more complex and somehow we did this through standardization and there's this interesting thing that's, like you said, we would say simultaneously, you are a veteran of the browser wars. We say the browser wars happened and we say ie6 won the browser wars, but at the same time, it's almost like all the people are for the same thing and they're really friendly too with each other.
Chris Wilson: Usually. Yeah.
Brian Kardell: I think there's a lot of comradery and encouragement among-
Chris Wilson: Absolutely. I mean, I think that the vast majority of people that I've worked with or worked in concert with, from other companies, even when we were competitors, we largely want to see the web move forward. That's the shared goal. I mean, that's why I moved from Microsoft to Google was because I felt like I could do more moving the web forward at Google than I could at Microsoft. And I think that there's a lot to this analogy of a browser war, because to me, one of the biggest challenges with this war analogy is you just conquer everything, And the problem is really, and this is what happened with ie6, what do you do once you've conquered everything?
Brian Kardell: And as I said, people are generally friendly and working toward the same things. I'm wondering are there things that we learned along the way that have helped us get better at avoiding the winner-loser scenarios? And to this point I ask, there's an interesting tradition in browsers that involves cakes.
Chris Wilson: Yeah.
Brian Kardell: And I'm wondering if you could tell me a little bit about that and maybe that's the answer is we just send cakes.
Chris Wilson: Yeah. Just send cakes, send cake, it can't go wrong. I mean, I think from very early days, there was a lot of, let's say playfulness about the competition. I remember, I think it was ie3 release party. They held a release event. It might have been ie4, but I think it was ie3. They held a release event down in Mountain View or the bay area somewhere, and as they had this giant ie logo that they'd created at the launch event and some of the leadership of the team who were down there at the event, I was just, lowly engineered on the team thought it would be funny, they packed up this giant logo, eight feet, 10 feet tall. I don't know how big it was, but it was bigger than a person. They loaded it onto the back of a flatbed in the middle of the night and took it over and dumped it on Mozilla's doorstep.And then there was all the news stories the next day. Mozilla actually took it in good fun. It was funny. But this turned into, I think this tradition of sending cakes every time a browser did a new release and we got a cake when ie finally delivered ie7, we got a cake. The ie team started sending cakes to Firefox and the like, and obviously at some point, this broke down. I mean, we ship a new version of Chrome once every few weeks now it feels like.
Brian Kardell: Chris, it's always fun to talk to you. I really like these topics and hopefully we can do more of them. It's always fun. Thanks for joining me.
Chris Wilson: Likewise. Thanks so much for having me.
Tab Atkins
Brian Kardell
Chris Wilson