Open Enterprise

RSSSubscribe to this blog
About Author

Glyn Moody's look at all levels of the enterprise open source stack. The blog will look at the organisations that are embracing open source, old and new alike (start-ups welcome), and the communities of users and developers that have formed around them (or not, as the case may be).

Contact Author

Email Glyn

Twitter Profile

Linked-in Profile


Mozilla's Brendan Eich on JavaScript - and Microsoft Buying Netscape

Article comments

It seems so long ago now, but for those of us lucky enough (and old enough) to have been there, the launch of Netscape's 0.9 version of its Netscape Navigator browser in October 1994 was clearly the beginning of a new era. For a few years, Netscape was the centre of the Internet universe - it's home page was the first you checked each morning for news about what was happening on this strange new Web thing that the company was doing so much to define.

Of course, Netscape went from Net hero to zero remarkably rapidly, but its influence is still felt in multiple ways. Perhaps most importantly, Netscape Navigator gave rise to Mozilla and thus Firefox. It also gave us JavaScript. Or rather, Netscape's Brendan Eich gave us JavaScript. Eich is not so well known as some of his colleagues around that time, and that's a pity, because he was one of the key figures in this formative time. Today he is CTO of Mozilla.

A few weeks ago I met up with him, and we looked back at the amazing ride the Web - and he - has had since the founding of Netscape. What follows is the first part of that interview, dealing with the birth of JavaScript and the origins of Mozilla. Next week the story will look at Firefox and more recent developments.

Inevitably, the first thing I asked about was JavaScript - and how it had come about. "It was certainly rushed," Eich says. Originally it was called Mocha, then LiveScript and finally became JavaScript once Sun had agreed Netscape could use "Java" in the name. Actually, it was Bill Joy, one of Sun's founders, who agreed to that, and Eich says that Joy's "name was mud at Sun the next day", because he didn't exactly have everyone's agreement on the move.

Indeed, Eich admits he "would have preferred a different name", but it was chosen because it was viewed as a companion language to Java. "Microsoft had C++, and had Visual Basic as scripting language companion," Eich explains. Netscape wanted a similar pairing. "The real push for JavaScript was to make a language for designers and programmers who wanted to mingle HTML and JavaScript."

Initially, Eich went to work in Netscape's server group. He explains: "I worked there for almost a month, in May switched to client group, spent 10 days prototyping the core language. When Marc [Andreessen]" - one of the creators of the original Mosaic and co-founder of Netscape - "saw that [prototype], he said: this is it, it's Emacs, we're done, and then immediately pushed to get it shipped."

Getting JavaScript out early was important, because Microsoft was "coming after Netscape," Eich notes. The relationship with Microsoft nearly turned out rather differently: "they'd made an offer to acquire Netscape late 1994 before I joined," Eich explains; "it was not enough money so they were turned away."

So far as I know, that's the first time anyone has mentioned this rather significant fact. After all, imagine what would have happened had Microsoft raised its offer, and Netscape had been acquired at the end of 1994, just after Netscape Navigator came out. That would have meant no browser wars, and hence Microsoft's domination of the Web browser sector would have been complete right from the start. Indeed, the Web might simply have turned into a Microsoft-only zone, since it would have been able to lock its browser into its other products without any serious challenge.

And no Netscape means no Mozilla, and no Firefox (although other open source projects would surely have sprung up to fill that gap.) It's a pretty frightening prospect of what might have been.

The threat of Microsoft's VBScript encouraged Netscape to take JavaScript to a standards body, ECMA (the official name of the standardised language is ECMAScript.) "We standardised it and that seemed to do right by the Web," Eich explains. "But Netscape was not doing right by the Web in general." This was during the browser wars, when both Netscape and Microsoft were adding non-standard features to successive versions of their browsers, essentially dividing the Web into two incompatible spheres.

"Netscape was under-investing in its browser, because it was using its IPO money to acquire a bunch of other companies," Eich notes. This was part of the management's attempt to turn Netscape into a general enterprise software powerhouse. The trouble was, they were losing their core market - the browser - and with it their one advantage over Microsoft. To make things worse, "Microsoft was making the browser into a platform" - a very shrewd move.

"We could see the market share dropping form peak of about 80%. It was clear that the browser was going to die. Management wanted to do something with the browser that would preserve it for posterity and possibly give it a second life," Eich recalls, "and that was the Mozilla project."

"We started in late 97," Eich explains, "I was ready for something new." At that point free software - "open source" hadn’t yet been coined as a term - was by no means mainstream, but the people starting to gravitate towards the Mozilla project knew that it was going to be big.

"those of us that dug into it realised this was important to do right," Eich says. "We took it seriously because if Netscape was doing this, it's a big breakthrough for open source, it's the first major commercial release. So we had to set up a virtual organisation that acted as the representative for the rest of the world to get them to collaborate with Netscape so that Netscape programmers didn't get advantaged. As 1998 wore on, we open sourced it in 3 31 [March 31]; in October I wrote a new roadmap: we should clean the slate."

Unfortunately, by this time Netscape was in serious trouble. "They rushed Netscape 6 to market," Eich recalls, "Mitchell [Baker]" - now Mozilla's "Chief Lizard Wrangler" - and I advised them not to." Unfortunately, they didn't listen, and it was immediately evident from the parlous state of the Netscape Navigator 6 product that Netscape was doomed.

But as Eich explains: "the Netscape 6 experience helped Mitchell and I take control of the project, because [Netscape 6] was such a failure in the market. We were able to say look, Netscape has been hiring badly, been giving CVS [software revision control system] commit rights to whoever they hire. So we should level that playing field [by doing the same for volunteers]. So we levelled the playing field, and that hurt some Netscape people. I think it may have led to Mitchell being fired in 2001."

Eich points out that this produced a rather strange situation. "The funny thing is with open source, at that point you had volunteers, Mitchell had enough severance [pay] that she kept working the whole time, so the management that fired her was flummoxed, saying: I thought we got rid of her. We kept working as a virtual organisation."

Well, not completely virtual: there was one issue that needed resolving. "What are we going to do about the machines? Because we had not just Netscape/AOL infrastucture - a lot of Unix machines and a grabbag of PCs and old Macs running automated builds and so forth - but also contributed machines we provided the power for. So we had somebody in our virtual organisation who promised he'd keep the machines working or take the volunteers' machines out and give them back to us if we had to."

But that didn't prove necessary. "Fortunately we didn't, because Netscape knew that Mozilla was still vital to them. The rank and file engineers liked us and wanted to side with us. The builds we produced, supposedly for testing and not crammed with AOL features, were actually more popular than the Netscape official builds" - a sure sign of which way the wind was blowing.

"Originally Jamie [Zawinski]" - one of the key early figures on Mozilla - "said: nobody should get a binary from us, they should only get source from us. But pretty soon we said: look, how are people going to test? It's just a massive waste of power and cycles to make everyone recompile to get the same bits. So we ended up shipping these testing builds, and then we started thinking about usability. We never really got it right with the Mozilla suite which was the open source program you could build, but we did with Firefox."

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Email this to a friend

* indicates mandatory field






ComputerWorldUK Webcast

ComputerworldUK
Share
x
Open