Why Worry About IE
Some Arguments Against Internet Explorer
by Charlie Loyd; last updated 20 September 2004.

This was written very quickly and therefore wordily. Feel free to skim.

There are many and many kinds of reasons to avoid IE. It’s a Microsoft product, and for some that’s reason enough to hate it, but we ought to look a little further and ask why it matters who coded it and what in particular could make it worth hating. This page is a polemical and rambling exploration of reasons IE sucks.

Microsoft is a monopoly. Even they would not dispute that. The chain of lawsuits against it allege not monopoly but abuse of monopoly power – everyone knows they control the markets, and the question whether they do it well.

They do not. Their strategy is world domination. World domination has become a catchphrase in this decade, a sort of charming token of ambition and irony, but Microsoft means it. Increasingly, Microsoft is so rich – and the legal punishments so small – that it feels no retribution.

Microsoft is playing the system. It can get away with practically anything because it can pay practically any fine, because millions of people and businesses rely on it (making governments hesitant to hurt it), and because it controls almost every market it’s in. With its money, connections, and lawyers, Microsoft is freer and more infuential than several nuclear powers. Truly the age of cyberpunk has come.

Anyway, Microsoft did not invent the internet or the web, nor does it help improve them. This is obvious to some people and surprising to most. Microsoft noticed the web rather late, and it’s reasonable to guess that they only got involved because they saw someone else trying to establish a monopoly. It was Netscape, who produced a rather good browser called Navigator that browsed (try to keep up with the jargon here) the web. Navigator was quite popular and its market was growing quickly, so Microsoft came along and released its own browser. Everyone tried it, of course, and now nine out of ten people who use the web do so with IE.

A loss leader, in retail, is an underpriced product advertised to bring people into the store. The company loses money on it, but it brings enough attention to other products to make a net profit. Microsoft used IE as a loss leader. People downloaded it because:

  1. It was free. Netscape couldn’t afford to make Navigator entirely free; Microsoft forced them to, but they never made much money after that, and now they’re practically out of business. To Microsoft, on the other hand, the resources needed to write IE were tiny – negligible – especially because they licensed the same foundation code that Netscape did. The web browser market was just spice on the entire software industry.

  2. It was from Microsoft. When people used IE, they knew they had Microsoft on their side. Most home users already relied on Microsoft and understandably favored it. You can hear in the way ordinary folks talk that they don’t distinguish between the act of manipulating digital information and the act of being a Microsoft customer. They know that Microsoft is big and important, and they’ll take what it gives. Microsoft’s ability to give them usable, capable, stable, secure software is secondary. More than someone to trust they need someone to blame. Everyone knows you can’t blame someone when Windows or Word or PowerPoint eats something they were responsible for, and to many that indemity from blame is more important than indemity from bad software. They would buy anything with more than 70% marketshare because it’s socially safe. The expression and organization of intellect and creativity doesn’t enter into their purchasing decisions. There are a lot of them.

  3. It didn’t suck too bad. IE did most of what it was supposed to, or at least when it failed it was pretty good at making it look like someone else’s fault. Bugs could be passed off as strange but conceivable design descisions. This was easier in the early days, when there wasn’t much agreement on how to make a given lump of HTML look to the user. For instance, should paragraphs be indented, or should they be separated by a blank line? Each browser was free to do as it pleased.

The leniency of of eo-HTML was too easy to exploit. Microsoft invented new tags out of whole cloth and made its browser work with them. It was sort of as though Nike were putting up big housing developments with sidewalks that only its shoes would stick to. It wasn’t that bad at first because other browsers could just ignore those tags and their users wouldn’t miss too much. It got a lot of people mad, though, because they saw the strategy behind the tactic: they saw that Microsoft was trying to entrench its users and slowly control the web by controlling the ways you could and should write web pages. Introducing a particular tag wasn’t an outrage – the stomping on the natural diversity of the web was.

(Indirectly, this abuse of HTML led to the stricter standardization of CSS and XHTML, and thus to better compatibility. But thanking Microsoft for that is like thanking a case of Hantavirus for reminding you to stay clear of rodent feces.)

But wait, what’s the harm in Microsoft controlling the web? They probably wouldn’t censor or anything, because that would be unpopular. Realistically, it’s not as though they would change much at all in an obvious way. Just as we submit to the largely benevolent control of a government, one might say, why don’t we allow Microsoft to encompass those parts of our lives we spend at computers?

First because it’s in our interest to have a choice. If everyone today chose Microsoft whenever they could (instead of 9/10 of people choosing it 9/10 of the time, as they do), in twenty years it would be impossible to choose anything but Microsoft. It would have an absolute monopoly rather than the ordinary monopoly it has now. Today, you can choose your operating system and your browser and so on, and even if you effectively don’t, it’s to your benefit that you could for two reasons:

  1. Even if you don’t pick one of the alternatives to Microsoft’s world, others can. Our society needs a huge range of software. For instance, servers tend to run Linux or BSD, and even if you don’t chose them for your desktop, the stability and flexibility of those servers helps you in many ways (your mail gets delivered rather than eaten, your website is easy to set up, projects like wikis can work, etc.). In the same way, people who write or work with graphics, sound, and video – people who use computers heavily, but not to make software – tend to use Macintoshes, even if they cost a fifth again as much as PCs, because to them the elegance and depth of the operating system itself seems more important than compatibility with the majority of home users. People who need these relatively unpopular things have to stick together to keep their suppliers in business.

  2. By having a choice, you implicitly threaten Microsoft’s power. You may not see it in terms of threats and power, but Microsoft sure does. If Microsoft knows that you don’t have to choose its products, it has to make them at least a little bit good. In the same way, in countries with one-party systems, the government is invariably corrupt, but if there’s even a small opposition party, they have to make a show of being worth voting for. As long as Microsoft has to compare itself to anything else – to say that it’s better in some way than its competition, and thus to admit that there’s competition – there will be a bit of perspective. People will know that there is an outside, even if they don’t go there.

So the opportunity to step outside Microsoft’s world makes your life better even if you never take it. In their world, you own a PC, you run Windows, you watch your movies in Windows Media Player, you read your mail with Outlook, you browse the web with Internet Explorer, you do your writing in Word, you chat in Messenger, you keep track of your money in Excel, you look for things with MSN Search, and so on. Every piece of software you use is from Microsoft.

One of the problems with the world is that it’s too smooth. Windows is famous for its viruses, but it doesn’t actually have that many security holes. The problem is that when there’s a hole in anything, it’s a hole in everything. If you open an infected attachment in your Microsoft e-mail reader, it can spread directly to the rest of the Microsoft system. As any ecologist will tell you, monocultures are vulnerable to disease. That’s why viruses are so bad these days – because most computers on the net are running Windows, and because most of the major software on them is from Microsoft. It’s tremendously vulnerable by nature. Breaking just a few links in the chain, replacing bits and pieces of Microsoft’s world by making routers run Linux and by replacing some of your Microsoft software with alternatives, strengthens the internet and strengthens your computer. They deliberately make their products work worse with non-Microsoft products, but you can do it, and I think it might be worth it. I’ve seen just a few stands of maple save a fir forest from laminated root rot.

Another problem with their world is that they control it. They own the search engines people use to find facts, the news feeds that supply information, and the browser people use to get to them. The more attention you give them, the more control they’ll take. Computers are tremendously powerful processors of information, and the internet is a phenomenally wide and influential means of communication; when you use them only by Microsoft’s permission, using Microsoft’s tools, you’ve lost something. Even if Microsoft never abused their power, I would feel nervous if everyone lived their way. It wouldn’t be a small system, but it would be a closed one.

To use a Microsoft product at all is to accept Microsoft’s rules of the game. They really do make their software work worse with stuff they didn’t write: that’s what they were doing with the HTML tags. Word, for instance, saves its files in .doc format, which is an intentionally obfuscated way of representing text and formatting. They’re making things painful for you if you want use something other than Word to see what you wrote in Word, and they know it. This kind of thinking is terribly counterproductive to users, and it’s extremely influential. Thousands of developers take their lead from Microsoft, and hundreds of thousands of tech support calls are made by people who can’t find the right program to open a file. And as with file formats, so with everything.

All their code – the stuff used to generate the programs – is secret, so there’s nothing you can do if they don’t do what you want. It’s like cigarette manufacturers not warning people what-all they’re smoking. It’s quite a lot like food, actually: the source is like the ingredients, and the software is like the meal. Microsoft sells you bread without telling you the ingredients, and if you want slightly different bread, and go bake your own, they can sue you for violating their trademarks, patents, and copyrights. You have to take Microsoft’s stuff or leave it; you’re not allowed to tweak it.

The big alternative, of course, is open-source software. Open-source projects, like the Linux kernel with the Gnu operating system, the BSD operating systems, the Gnome desktop, the Apache web server, the Mozilla browsers, and the X-Chat IRC client, don’t have any secrets. Their source is public – you can download it for free and change anything you don’t like about the program. You can even e-mail your changes back in, and the volunteers who maintain the projects will consider using them in the main version. They’re trying to make good tools.

Open-source software has its bad points. It’s often harder to learn (though not necessarily to use) for the average person, because it’s mostly designed by programmers, not designers. There’s usually no central help-line like you have with Microsoft; you often have to ask a volunteer on the internet if you need help. People like me, who have made some effort to learn the ways of computers, forget how difficult it is for just plain folks to use the things we consider idiomatic and clean. The famous annoying answer to requests for help on something open-source – the equivalent of please hold for closed-source software – is read the source. The implication, of course, is that the listener is a competent programmer and will understand the program by reading the C or C++ or Python or Perl or Ruby or Lisp or whatever language it’s written in. Not all that many people have bothered to learn, so you can be almost as stuck as though it were secret.

The heartbreak, for me, is what so many people have bothered to learn. Figuring out how to use Windows is harder than learning to read the source of an average program. This is rather clear in my mind, as in the last year I have for the first time been forced to work with Windows, and it makes me angry. It amazes me that average people all over the world have the patience to put up with this every day just to get their work done. It infuriates me that they have to: that they’re putting up with this rather than with an equally annoying but fundamentally better way of using computers. It’s a great deal of wasted effort. I feel terrible when I imagine the people who are permanently soured on all the beautiful things computers can show just because Windows always annoyed them. I can’t blame them and I wish I could help them. Windows presents so much of what people complain about when they complain about computers – dehumanization, arbitraryness, needless complexity, and confusion – that I think we’d be a noticeably healthier society, and more reconciled to our tools, if it hadn’t become the default.

Rather than get angry at it every time there’s a good reason to, I try to ignore Microsoft. I can’t, because I do web design, and most people see the web through Internet Explorer. About a quarter of my work time is spent coding, a quarter constructing and editing images, a quarter writing XHTML and CSS, and a quarter figuring out why it doesn’t work in IE. That’s a lot of time. IE is a buggy mess – a moment’s Google will turn up dozens of open holes in it.

It works, of course, enough that people don’t complain much. But its most fundamental problem, the thing it’s hiding, isn’t philosophical or abstract at all: it renders stuff wrong. Given correct XHTML and CSS, it gives you something else. It makes it look like the person who wrote the web page misdesigned something when no such thing is true. XHTML and CSS are fully standardized – there are clear instructions for how you may and may not treat it – and IE breaks the standards all over. Microsoft claims to aspire to standards-compliance (they certainly have representatives on the standards committees, many of them wonderful as individuals) and they have a huge organization of coders, but they can’t seem to do it right. Microsoft hires some of the finest programmers alive today, and they turn out lousy software. Their collection of coders is a once-in-history kind of thing, like the Roman engineers or the Babylonian gardeners or the Dutch painters or the WWII Allies’ mathematicians, but somehow they just can’t cut it. IE is not as good a web browser as its competitors. IE is less secure, doesn’t render stuff as well, doesn’t have tabs, isn’t updated as often, and so on and on. There’s nothing that makes it better.

Well, one thing: inertia. The inconvenience of (a) learning that IE is not the web itself and (b) downloading and installing Firefox after having set up some bookmarks and so on in IE keeps millions of people from using better software. That’s it. That tiny, five-minute trouble – and the face of unquestionable authority that Microsoft shows its customers – is all that gives IE 92% market share.

To me, as a web designer and someone with an annoying urge to use good tools, whether they’re pens or bicycles or hatchets or paintbrushes or eating utensils or washing machines, IE is poison. I handle it only to make sure that its users aren’t hurt more than they have to be. After using Mozilla Firefox or Safari, web browsing with IE is like eating a good meal with a plastic spoon. Sure, it’s enough: just enough to keep most people from realizing that they could ask more of a browser. The web is a fraction uglier and harder to use for them, and they’re either okay with that or they don’t know it could be different. What itches me is that they never have a chance to choose IE; it’s chosen for them and they don’t even know it.

One more point – it’s not just a matter of taste to use Safari or Firefox; it actually indirectly and slowly improves the web for everyone.

When your browser requests a web page from a server, it sends messages called headers. The main one names the page it would like to see. Others define the preferred language, for instance, so a web page that comes in several languages can decide which to use. The user agent string tells the web server what browser it is – its name, version, and so on. This lets web pages come in different versions to correct for IE’s bugs, for instance. Servers record headers to keep track of what files get requested when, by whom, and how, and part of that is the user agent.

Webmasters analyze these logs to see what pages are most popular, to produce hit counters, and especially to see which browsers people are using. Web designers keep track of these statistics, and the more mercenary and parasitic of them design strictly for the leading browser. Right now that’s IE 6, so thousands of people are doing whatever it likes and completely ignoring the standards. The more responsible designers – those few – design first for the standards, and then, holding on to that, skootch around until IE doesn’t blow a fuse. We should cultivate them.

Please choose your browser carefully. Thank you.