The end of the one-browser Web

June 3, 2002

By the time you read this, the final version of Mozilla 1.0 might have been released. Or maybe not. It’s been a long time coming, and there were times during its long gestation when it seemed that it wouldn’t be worth the wait. Now that it’s here, there will no doubt be dozens of articles pontificating about its implications for Web authors and Web users. This is one of those articles.

Firstly, Mozilla will change the browser profile of your audience. We already know that AOL has been beta-testing Mozilla’s Gecko rendering engine as the core of version 7 of the AOL and CompuServe client packages. Over the years, people have joined the dots between AOL, its subsidiary Netscape and the Mozilla project, and have wondered when this would happen, but I suspect deep down we didn’t believe it would. We’re so used to the usual deer-in-headlights pose that other IT companies adopt when dealing with the Microsoft monopoly that we’re surprised when someone actually competes with them. It will take some time for the users to adopt the new browser; if our stats at MacEdition are anything to go by, there are still a few AOL users using version 4.0 of their client software, so we shouldn’t assume that all AOL subscribers will upgrade at once. But upgrade they will.

I think we will also see a fair few Netscape 4 users upgrade, too. People often attribute the continued usage of Netscape 4 to user ignorance or conservatism, but one of the surprising things about MacEdition’s traffic in recent months is that Mozilla has been outstripping Netscape 4. Despite my earlier optimism, Netscape 6 usage was never very high – a bit higher than Opera or iCab, and nowhere near OmniWeb or Netscape 4. But Netscape 6 and Mozilla combined still account for around 10 percent of our traffic. MacEdition probably attracts an audience that is more likely than average to use the latest and greatest browsers, so it’s a bit hard to make too much of this trend for other sites. Nonetheless, since the beginning of the year, we’ve seen Netscape 4.x usage continue to drop, and Mozilla’s has risen. At the very least, people upgrading to Mac OS X will stop using Netscape 4, as will people upgrading to new computers. Jakob Nielsen’s prediction was that we would be stuck with version 4.0 browsers until 2003. It seemed appalling at the time, but it’s now only six months away. (Even more appalling for some people – having to admit Jakob Nielsen was right!)

Secondly, as the market share of the Mozilla family increases, Web authors will find that the old comfortable habits from the days of Netscape 3.0 won’t work any more. Improperly nested code will come out badly, and even some valid markup will look different to what you expected. Fortunately the long gestation period of the Mozilla family means that the more forward-thinking Web authors have had plenty of time to experiment with what works and what doesn’t. There are general guides on Mozilla-friendly authoring, and specific explanations of how some interpretations of Web standards differ from others. It’s an innovating Web out there, so it’s time to Upgrade Your Brain and start thinking about standards and accessibility.

Let a thousand lizards spawn

Because it’s an open source project, Mozilla has generated several side products that will also contribute to the standards-compliant Web. Netscape 6 and now 7 are the obvious examples of this. The speed and stability of Netscape 7 is so much better than the preview releases of Netscape 6 – nay, even most of the release versions – that it will probably grab a fair bit of market share. I wasn’t too surprised at the improved performance, having had Mozilla 1.0 release candidate builds in my testing armoury for some weeks. Still, I was struck with the contrast from Netscape 6’s preview releases. Back then, I was flaming Netscape for putting out such resource-hogging, steaming piles of instability. Now, despite a little shakiness typing into text areas, I’m able to use Mozilla RC2 or Netscape 7 PR1 as my daily-use browser, although as I mentioned in a previous column, I still find myself going back to IE for some tasks.

For Mac OS X users, there’s also Chimera, which has made inroads into shares of other OS X browsers despite its obviously unfinished state. The fact that the rendering engine is open source and easily licenced means that we should expect even more such Mozilla offshoots in coming months.

Incidentally, this is where Vincent Flanders has it all wrong. He wants one browser, if only half seriously. The problem with this thinking (aside from the fact that we know whose it will be) is that the whole package shouldn’t need to be the same. People want different features. If I don’t want IRC clients and Net radio and stupid IM clients that probably don’t work inside our firewall (and like I could deal with more interruptions than I get already anyway) – then I don’t have to have them. If what I want is just a browser, I can have it. If I’m wedded to Entourage but not IE, I can have Chimera. If I’m hooked on graphical editing of Web pages, but prefer Eudora to Mozilla’s mail client, then a simple installation option (currently for everyone but OS X users) fulfills that need, too. If I want something that didn’t look like a Linux nerd designed the interface, the Goddess might be the answer. In principle, I could build such a browser myself using XUL. We could have many browsers, one rendering engine. Or we could have something like what the market seems to be evolving towards – two or three families of browsers, each built on a core of standards-compliant rendering.

Yes, there are glitches and differences between, say, IE5/Mac, Opera and the Mozilla family. They’re not the big deal that we used to encounter in making things work similarly in the various Version 4.0 browsers. Isn’t this what we have been hoping for all along? We don’t need a single-browser world, any more than we want a world with just one operating system. Compatibility is about file formats and communications protocols, not the viewing applications or the infrastructure around those viewers.

Something for my work chums

The final point I want to make about the implications of the Mozilla project is nothing really to do with the browser. Everybody looks at the browser. What is less recognised is that Mozilla and its Netscape sibling both come with a decent, standards-compliant and graphical editor for Web pages. The importance of this has not been widely recognised. People think of Web pages, Web sites, being made by professionals – bearded coders and black-clad designers. The non-coding, non-Photoshop-using unwashed are presumed to be tooling away in Blogger or something similar, wishing each other happy birthday, taking sides in the Middle East conflict, or posting pictures of their cats. This is a misconception. There are huge numbers of “non-professional” people who need to create documents that display on the Web. They don’t want to learn HTML if they can avoid it, and they definitely don’t want to care about what works in Netscape 4, or the differences between different versions of IE in rendering absolutely positioned boxes.

Mozilla Composer, or some evolution of it, is what thousands of college professors need to produce the course information they post for their students. It’s what the UN needs to produce its media releases instead of generating multiple kilobytes of steaming Microsoft-proprietary pseudo-XML bumf for a two-paragraph document. And if its developers get their act together and make it easier to user templates and link to existing stylesheets, perhaps using the CaScadeS project, it might even be the right tool for the thousands upon thousands of people who create content for their organisations’ intranets.

Oh yes, Mozilla matters, but not necessarily in the ways you might think.

— CodeBitch (codebitch@macedition.com) is the grumpy cow who does the HTML production for MacEdition. Read other articles by CodeBitch, or check out my interview over at Lemurzone’s pixelview.

E-mail this story to a friend