Dodging the Draft of the Browser War by Mike Sugimoto, who doesn't believe in solutions

April 2002 Update: I take it all back, every single bad thing I say about Mozilla here. As I write this, Mozilla 1.0 RC1 is running happily on my browser, and it has consistently impressed me in its performance and abilities. If you haven't tried a Mozilla lately, you really owe it to yourself to go and take a look. The things I thought were stupid when I wrote this article -- skins, sidebars, tabbed browsing -- all of it has proved to be incredibly useful. I use the sidebar frequently. I love being able to do tabbed browsing (center-click will open it). These features are a lot like mice with wheels on them: You think it's a gimmick, until you actually use it, and then you can't live without it.

So congratulations to the Mozilla folks. You're ever so close to a version 1 release, and the current candidate rocks my computational world. Best browser out there, hands down.

The only point I'd like to make is that it's criminal it took two years since the time I originally wrote this document before we got a Web browser that really didn't suck. But if that was the price we had to pay, I think it was worth it.

It all started back in July of this year, down in Seattle. Frink, Axxia and I were at a couple of talks given by my personal Internet hero, Philip Greenspun, on his favorite topic -- Web design and applications. At one point during the presentation, Philip began to curse at the laptop. "Netscape sucks," he said. "Internet Explorer is a much better browser. We'll have an all IE afternoon," he promised. The rotation of my head towards Frink and Axxia (seated on opposite sides of me) probably induced some significant centripital force on the critters that live in my hair.

My confusion might have been understandable. Philip Greenspun praising a Microsoft product? Isn't this the guy who put together the Bill Gates Personal Wealth Clock and who savages Microsoft at every opportunity? Didn't I hear him get criticized for having an obvious anti-Microsoft bias during the first day of the talks? Granted, we were in Seattle (not exactly a hotbed of anti-Microsoft sentiment) so maybe he was moderating his views a bit, but then again, I've seen him rip on vendors that piss him off with little regard for whether or not they like him or not, and I've seen him say some pretty provocative things.

Maybe there was a kernel of truth to what he was saying.

It turns out that Internet Explorer is a much better product -- at least if you're using Windows. My experiences with Netscape were pleasant right up until they kicked everything up to Communicator v4, when everything became disgustingly bloated and slow. I recently tried to downgrade to a version 3-gold release of Navigator, and ended up discovering that Netscape's home page contains Java that will break the engine living inside verion 3 clients. To get full functionality out of most Web sites these days, it seems you need to have the latest browser from either Microsoft or Netscape; in practice, this means IE5 or Netscape 4.75.

I could have gone on using Netscape 4.7 very happily for a while, were it not for that nasty Brown Orifice exploit that came out in the middle of August. Upgrading to 4.75 was a near-painless experience, but then the crashes started: Netscape would hang at the slightest bit of weird code, and often for no good reason. It seemed to me, however, that most of the problems I was experiencing with Netscape involved pages that made extensive use of third-site images and cookies -- essentially, anything that used advertising, and thus most commerical sites were affected.

It's not clear exactly what happened to cause my browser to crash, but I'd have to kick it over by banging on the "End Task" button in the task manager, after reading the little "(Not Responding)" label next to the window title. Netscape also had an annoying habit of getting wedged so that clicking on hyperlinks wouldn't actually lead you anywhere (though right-clicking and opening into new windows would allow you to get what you wanted), and exiting the program didn't actually solve the problem. (Again, you had to enter task manager and manually kill the process.)

I had high hopes for Opera when it was first released, but its horrible Java and CSS support have kept me from playing with it too extensively. And I really don't like the nag ware feature, and the fact that it breaks after 30 days if you don't fork out $39 for it. Mozilla represented my last, great hope for a Netscape-ish browser, but as it stands, the project has several strikes against it:

• As it stands now, the binary releases are full of debugging code, making the software glacially slow. It feels like I'm doing X over a very slow network connection -- the windows lag behind my mouse inputs -- and this is on a P266.
• Mozilla's development team seems to have put an awful lot of effort into making their browser "pretty" rather than functional. I don't want a sidebar. I don't want themes. (If I wanted themes, I'd get a new window manager.) I don't want every element of my user interface to be defined as a bunch of .gif files whose status needs to be tested for stupid things like onMouseOver() to see whether a new file needs to be loaded. Web browsers do not need skins.
• There isn't nearly enough user level customization available.

Why Mozilla Just Doesn't Get It

• The Gecko rendering engine
• Internet Explorer's understanding of cascading style sheets (way better than Netscape's; this may live inside the rendering engine)
• Plug-ins that were disabled by default (i.e., you don't suddenly discover that it's taking Netscape a minute to load a plug-in to play a background .wav file some inconsiderate publisher thought would be cool to have)
• No default file associations, so that everything that isn't a .txt or .html file has the option of being downloaded
• No mail or news client
• A provision for disabling frames
• As much of the screen space as possible dedicated to the Web page itself.

In default operating mode, the left third of the screen is dedicated to "My Sidebar," a feature whose purpose I have not yet determined. Much is made of the fact that Mozilla supports skins -- a potentially interesting feature in WinAmp, but not something that I've taken advantage there, and not something I would ever take advantage of in a Web browser. I don't see the point. The developers need to have Alan Cooper's famous saying tattooed on their eyelids: "No matter how cool your interface, it would be better if there were less of it."

In a perfect world, there would be two versions that ultimately come out of the Mozilla project. One would be a full-featured bells and whistles browser for the general public, and the other would be a stripped down machine that goes like snot and doesn't do anything other than render HTML and display Web pages. If it's possible to do both right now, I haven't figured it out. I can't speak for my friends, but if I found out the Mozilla team had ripped out most of the guts of their browser and slapped on a minimalist user interface, you'd have to DoSing my connection to keep me from grabbing it. I suspect most other hard-core Web users feel the same way.

This could well become a rallying cry for the next generation of browsers: Less browser! More page! The browser should be nearly transparent to the user, if that's what the user wants -- almost the entirety of the screen should be dedicated to the display of the page, with minimal UI cues. I rarely need more than an hour to get to know a new program that does essentially the same thing as an old program; Web browsers are dead simple and shouldn't need any kind of a learning curve for anyone who has used one in the past.

Thinner browsers force developers to make decisions about bloat, and about what goes into their product. There is a disturbing trend in software development today towards bigger and bigger products; this is particularly prevalent in the Windows world where Microsoft seems to be the biggest offender. Although I'm not a software developer, I don't see the point in including the kitchen sink in a product that hardly has any use for plumbing, never mind culinary plumbing. Most Web browsers don't need a lot of UI; most of them don't need mail and news clients, and most of them don't need anything beyond a rendering engine and a way to get that on the screen, maybe a JVM. The rendering engine is small. Why is the resultant product so big?

If someone has an answer to this, please e-mail me. I'm looking for an answer.

So Why Not Give Up?

The Bottom Line

I never thought I'd be saying that about something that came out of Redmond, but there it is.

Update: 23 October, 2000

So why do I still not like Mozilla?

• IE-style right button behavior: IE needs you to let go of the mouse button before it pops the menu up; Netscape doesn't. I like Netscape's behavior more since I tend to use right-click-back instead of the back button on the menu bar. (User issue; not specifically Mozilla's fault.)
• Browser does not mark links as read. Blue links do not turn purple, etc. (Mozilla engine issue; big problem.)
• ^O is now "open file"; ^L is "open Web location." This is extremely annoying since ^O works to open both files and Web locations on Wintel versions of IE and NS; ^L is "open file/Web location" on the Mac, and the shift is annoying. (UI cuing problem.)
• Browser does not hold page position on re-entry through "back." IOW, if you follow a link half-way down a long page, then click back, you will be returned to the top of the page. Very annoying. (Mozilla engine issue, also a big problem.)

More

• Jamie Zawinski, one of the originals from Netscape and the one-time manager of the Mozilla project, has a few choice words of his own about the failure of Mozilla. (Excuse #2 is most damning, in my view.)
• You can presumably still get bloated Netscape software from their Web site, but you may have to dig around a bit.
• There is an extensive FAQ on Mozilla's interface design. (None of which makes any sense to me, but your mileage may vary.)
• An old, but interesting The Standard highlights some of the institutional schism involved in Mozilla.

Add/view comments