A Canadian who once bore the strange title of “”international incident”” says the release of Mozilla 1.0 represents a global step forward for open source developers everywhere.
Mike Shaver got involved with the Mozilla Project in the fall of 1997 while working as an engineer in Netscape’s Java
script team. He had been lobbying the company to licence the Java script engine under the LGPL when he learned Netscape was considering releasing its source code for developers to tweak and improve. This marked one of the most significant open source projects in the history of technology. Shaver succeeded Mozilla founder and open source advocate Jamie Zawinski, becoming head of developer relations for Mozilla.org. He left in 2000 to join Montreal’s Zero-Knowledge Systems as chief software officer on the condition that he could continue to work on the project part time.
Having since left Zero-Knowledge, Shaver now acts as an America Online Inc. consultant. He shared war stories with Computing Canada about the browser’s development and its potential in the enterprise.
CC: Netscape 6 (which contained Mozilla) was filled with bugs. Was anything learned from that launch that was a sort of guiding principle to this release?
MS: I don’t want to speak ill of Netscape 6, and that really does feel like a long time ago (laughs). Netscape was a very educational experience. It was the first product launch of that scale, based on our technology. Just going through that release cycle was very valuable. We got to see from various online reviews and user feedback from all sorts of platforms what pieces of user experiences they were happiest with, which ones they thought needed the most work. We certainly took that into account. The Netscape 6X release provided a lot of powerful feedback. We had a lot of wider desktop and consumer reach than our pre-releases.
I think everybody has something about Netscape 6 they wish they could have been fixed before it was out and I think the fact Mozilla 1.0 was quite a while after that indicates that there were things we were concerned about. The embedding story is a good example of that. Netscape 6 wasn’t put out with the intent of solving that embedding space problem, but that’s a very important part of what we want to provide in terms of technology.
CC: Why did it take four years?
MS: We made a lot of changes of the way we manage our testing, to the ownership of our major corporate sponsor. Then in the technology as well — in the fall of 1998 we threw out much of the code we had because we realized it sort of reached the end of its ability to be enhanced and do the things we wanted it to do. I think that event was sort of foreshadowing to the rest of the project in that it was a huge change in development and product strategy for Netscape. And it was largely brought on by feedback and reasoning and lobbying on the part of the Web development and open source community.
There’s been a lot more of that since. Certainly it hasn’t been all wonderful and positive. You can’t do four years of a project and have it all be sunshine and flowers. Mistakes were made, as it was said in the Iran-Contra affair. But I think on the whole, everybody is very happy with where it is. We’re certainly very happy with the state of Mozilla 1.10. It’s not perfect, and we’re hard at work now on the first preview release of 1.1, which we hope will be out, we hope, in the relatively near future. It took us quite a while to get 1.0 in the state we wanted — there were last-minute fixes trickling in and getting more and more broad testing as we went through the release candidate process. We’re still hard at work on that stuff. I think the relationship between open source and corporate development has changed a lot since 1998, and I think it changed a lot for the better. I like to think Mozilla had a positive influence on that, demonstrating one way, if not the only way, of corporate development moving out into the open. And I think still sort of unique in the scale of the project that a company took its flagship software and essentially overnight decided they were going to play on an even keel with all the open source developers that were going to work on the project.
CC: Milestones like this are important too, because they show that things can be accomplished, and it’s not always going to be at that early stage.
MS: Absolutely. There were a lot of people, some involved with the project, and I think everybody had their days when they thought we were never going to get to 1.0 (laughs). We hit a bad point where we took a bunch of progressive changes in a short period of time and we fell a long way from where we wanted to be performance-wise. It took a lot of effort to get back on that path. And that’s just a facet of software development. I think the fact that it was open made it easier to see that and see some of the seedy underside of that process.
CC: How do you think AOL is going to make use of it? I’ve heard it’s being tested in some versions of its product.
MS: They certainly have expressed that they want to get it into their client, and Compuserve is doing tests on it. There are a lot of organizations that want to embed some or all of the Mozilla in their project, and AOL is one of those. I mean, it’s a huge technology decision for them. Their product’s been built around Internet Explorer, embedding with it for a long time. It’s not a trivial change for them.
In one sense, our goal is that when they switch over, the user won’t know. It will be just as fast, it will be just as powerful. And I think we’re pretty close to that. The standards support should be better for them, some of the ability to integrate with the AOL content should work better for them. But I think just generally they’re interested in it for use in their client. I think they’re interested because it’s a key part of their Netscape product technology, and I think they’re interested in it because they, like a lot of organizations, see the value in a heterogeneous Web, where the technology and the standards aren’t just controlled by one organization.
Our job isn’t to beat IE. If IE were taken off the market tomorrow our job would become more important, not less. But I think they have a vested interest in having an alternative option. Having Microsoft control that technology puts them in a bad negotiating position and companies are all about avoiding such position.
CC: We’ve heard a lot about the potential for Mozilla in non-PC devices. There are already browsers like Handspring’s Razor and LinuxLab’s Vagabond for handhelds, but what about tablets and the like? Could this find a home there?
MS: We’re certainly interested in those spaces, and people in those spaces have been interested in us, from Tuxia and WorldGate doing set-top type boxes to organizations like OEOne, who are building an entire development environment. There is a lot of breadth there, in terms of the capabilities of the device and so forth. This is something where we sort of look to what Linux is doing, where it runs from literally a Palm Pilot to huge clusters.
We’d like to get there. It’s pretty hard to build technology, especially application technology, which is competitive with a big desktop browser, (something) that’s very feature-rich and able to take full advantage of the full capabilities of that system and runs on a PDA. We do it down pretty small — the set-top and kiosk community are helping us there and giving us good feedback. The embedding approach — where we take the core layout engine and networking technology but not the whole application UI and put that in other software — is working well. Our footprint’s coming down nicely. I think it’ll be a while before we run on your Palm, but I think it’s also the same sort of thing the other way around. The companies that have technology in that space, like Razor and whatnot, they don’t really scale up to the desktop either. And it’s just because it’s a hard problem. It’s not that one set of technologies is better than the others necessarily. We started to approach it largely because of our heritage from the desktop. That’s where the browser was. As small devices become more capable, and as we learn how to do more efficient Web rendering, we’re going to meet in the middle. And I think it’s going to be powerful when we do.