The Browser Wars

At some point, hopefully quite early in the process, your web designer/developer is going to broach the subject of ‘browser compatability’ with you.

The following is intended to help you do rather more than just nod your head sympathetically, and leave it all up to her.

One of the longest standing issues in web development is - do you strive to support every browser, most browsers or just the one which is currently winning the browser war.

At the moment, an increasing number of sites go no further than to support Microsoft's Internet Explorer, versions 5 & 5.5, on the basis that this covers 80-85% of all the ‘hits’ coming its way. Other brave souls go to extraordinary lengths to support nearly every browser on the market. There ought be a simple compromise - just adhere to standards and everything will be all right. Unfortunately, clear though the World-Wide Web Consortium's standards are, strict adherence to them would probably break the majority of web sessions, because none of the most popular browsers follow these standards.

That's the essence of the browser wars - Netscape and Microsoft tried to impose their standards on the WWW, inflicting damage on the medium which may take years to repair.

...Microsoft. At least, at the time of writing (mid 2001), that's true. It's worth noting, however, that Netscape was declared the outright winner, not so very long ago. Things have changed, and they'll change again. For one thing, Internet Explorer 6 seems set to be something of a new departure - it purports to be much more standards-compliant than any previous MS browser, yet some of its ‘extensions’ seem likely to stretch those standards to breaking point (again). Meanwhile, Netscape 6 is creeping over the horizon, also promising to be standards-compliant.

Once again, the loser is going to be the web designer/developer who wants to serve as many potential readers as possible, with the same set of code.

So you'll settle for 85%, right? There are 429M people on the Internet; 85% of that is a very big number (364,650,000). Surely, that’s plenty.

A lot of people are making that judgment. I'm not here to tell them they're wrong, exactly, but I do think this needs a bit of thought.

For a start, I'm more than a little suspicious of that 85% figure. Quite a few of the ‘fringe’ browsers actually lie to the web-server, pretending to be MS IE (this harks back to a time when a significant proportion of web servers would only admit ‘authorised’ agents). In addition, the statistics are skewed by the fact that they're self-selective; increasingly MS-friendly sites aren't visited by users of non-MS browsers, because the sites won't work properly for them.

But, what worries me much more about this 85% figure - is its worrying similarity to the odds of surviving a ‘game’ of Russian Roulette. I know the world of the Internet is supposed to be exciting, but I'd rather it wasn't that exciting.

Those are really pretty lousy odds - for anything that matters. Would a software company be happy to release an application which failed 15% of the time? Would a shop-keeper be happy if one in six of his prospective customers were turned away at the door by an over-enthusiastic doorman? How long would Budweiser survive if every six-pack included one can of vinegar?

Maybe, though, your website doesn't matter that much? Then why the hell are you going to the trouble of building one?

How would you like to get a call from Steve Case?

If you don't know - Steve Case is the CEO of AOL-Time Warner. If yours is a content site, or an e-commerce site, or a web service site, a call from Steve Case (or one of his minions) could change your life. Would you like to get a call from Steve Case?

OK, let's assume the answer is ‘YES! Please God. Let it happen to ME!’. Let's say Steve hears something nice about your company, and decides to take a look at your website. What browser do you think he'd be using? Hm?

Well, if you've done your research, you'll know that the AOL installation CD puts a branded version of MS IE5 on its members’ desktop - so there's a fair chance that Steve would be using that, right?

But, who owns Netscape? AOL-Time Warner owns Netscape. And AOL-Time Warner are known to be less than happy to be dependent on Microsoft for access to their customers. Maybe Steve would be using his own company's browser - if only to make a point.

But AOL is also involved in the development of something called Mozilla - an open-source browser which is destined to be the foundation of Netscape 6. That's a candidate, too. As is an early version of IE6, which might end up being AOL’s preferred browser if Netscape 6 isn't ready in time.

And Steve is reckoned to be a pretty 21st Century kinda guy - so he might be trying out the new Sharp Zaurus, which runs an Opera browser under Linux, or Konqueror under KDE/Linux, or....

The truth is, neither you nor I know what browser Steve Case uses.

It doesn't actually matter which browser is used by the majority of web users. What matters is - just how many people can you afford to piss off? And, can you afford to piss off Steve Case, if he drops in?

This is particularly relevant to business-to-business sites. In many of these, only a few thousand people, in the entire world, are in a position to put serious business your way (or a few hundred, or a few dozen). How many of them are you prepared to piss off?

This is one of the most common mistakes made by businesses building a website; if it works on the browsers we use - then it must be OK.
WRONG!

For a start, IE5 is lousy test-bed for checking a website (and it's usually IE5). It’s far too fault-tolerant - it let’s too much bad code go unpunished.

A page which looks perfectly fine in IE5 may look like a kidnapper’s ransom note in Netscape or Opera.

Resist the temptation to say: "Why don't they use the same browser as everybody else?". That's not your business. If your pages appear jumbled - or some of it doesn't appear at all - that's your fault.

If the paragraphs above have not convinced you that not all browsers are alike, if you're pretty sure that ‘everybody in our business uses Microsoft’ - let's try something:
Go to the View menu, select Text Size and click on Largest.

Whoa! You didn't know that was there, did you?

Now, on that same menu-item, select Smallest. Makes a difference, doesn't it?

Those choices are there for a reason - people need them and people use them. Some people’s eyesight isn't very good, so they need large text. Others don't like to sit too close to a screen, so they set up a big monitor, five feet away from their desk, and use a larger text setting.

On the other hand, some people use small screens - on sub-notebooks and PDAs - and want to cram as much information on to them as they can manage. They use smaller text settings.

People use their browsers differently - even the same browser. Some people like to use their browser Full Screen (try it - press F11). Others like to keep half an eye on their desktop, or on other applications, and shrink the browser window to a half its full size (or less).

In fact, if you really wanted to, you could change the look of this page utterly (using Tools, Internet Options, General, Accessibility). You could change the font and the font-size, the background colour and the foreground colours. You could switch off all graphics, freeze any animations. You could change the behaviour of any or all of the styles I've used, to the extent that I wouldn't recognise my own work. The only thing you can't do is change the content.

That’s what’s so special about the Web - why it’s different from any other medium that’s gone before. On the Web, the publisher does not control the means of reproduction. The reader does.

If so much variation can be created in a single browser, what about some of the people who might be browsing?

The Internet has been a liberation to many disabled people, who can compete on equal terms in this medium, and many businesses have benefited from these additions to their workforces.

Technology exists to give near-equal access to the Internet’s goodies. Do your pages help or hinder that technology?

Now, you may feel that catering to these minorities is a bit too much too ask. "We'd like to help, but...", "Our core audience isn't...", "Business doesn't have time to be PC..."

Let's be blunt for a moment; the days are gone when a blind/disabled man or woman was only employed out of pity - in lowly, non-decision making roles. Businesses need skilled labour at every level and there is no real reason why the CEO of a company you want to do business with must be like you. He/she may get about in a Bentley, a wheelchair, or with a white stick.

Further, there are now regulatory issues to consider. The U.S. ‘Section 508’* now requires companies who want to do business with the federal government to follow accessibility guidelines - and that includes web accessibility. Similar regulations will apply in the EU. In the UK, the Disability Discrimination Act is already law.

So, having implanted FUD (Fear, Uncertainty, Doubt) in your mind, how do I propose to resolve all the problems I've identified, and leave you feeling warm and safe?

Can't do it, I'm afraid. There isn't a simple answer. When your web designer comes to you for guidance on how compatible you want your web site to be, you'll still have decisions to make. There are coding tricks to recognise incoming browsers, and deliver appropriate pages accordingly - but building all those extra pages is going to cost money, and they won't cope with the different ways in which readers use those browsers.

You will probably have to decide that you really can't please everybody.

But, there is an underlying truth about the Web, established in its infancy but often forgotten since; the Web is about content. It isn’t a layout medium, it isn’t a broadcast medium. It’s a reader’s medium.

Get your content right, first. Add frills and frivolities if you feel you need them - but be prepared to lose them - if the reader’s browser (or the reader) can't handle them. Don't rely on frills and frivolities to make your point.

Design a layout which works in IE5, but "degrades gracefully" in other browsers (you can assume that a particular reader will use his/her browser consistently - all you have to do is present them with a site which behaves consistently).

Test your whole site in a range of browsers. If you only have time for one quick test (when updating a single page, for instance) use something like Opera - the most standards-compliant browser around at the moment.

For big sites, there are on-line services like Netmechanic which will show you what your site looks like in a range of browsers.

Pay attention to feedback from readers - it’s a good idea to have feedback mail link on your Home page for this precise purpose. If they’re having trouble with an aspect of your site, see what you can do to fix it (there are some people who would complain about the tiniest thing, but, mostly, they’re just doing you a favour).

Above all, put yourself in the shoes of the reader; for all the big numbers involved in the Internet, there's really only one reader at a time. Is he/she really going to be impressed with a sharp piece of layout - enough to pay out money/sign a contract? At the same time, is he/she going to be impressed if your page-layout falls to pieces in the browser they happen to be using?

You need them more than they need you.