This site’s design is only visible in a graphical browser that supports web standards (yours does not), but its content is accessible to any browser or Internet device.

RedGrittyBrick

Standards

Web Browsers

As developers bring out new versions of web browser, they are increasingly supporting the W3C standards. This should make it easier to design web pages that appear as intended in a variety of web browsers.

Internet Explorer 5 has some defects in the way it handles cascading style sheets (CSS). In particular, it calculates element widths incorrectly (i.e. not according to the spec).

IE6 handles this aspect correctly (i.e. differently than IE5).

Netscape Navigator 4 barely supports the CSS standards at all.

Netscape Navigator 6 supports CSS quite well compared to its predecessors.

There's a whole bunch of other browsers that may be increasing in importance: Pocket IE, Opera, Mozilla and many others. Some people also rely on plain-text browsers such as Lynx or use voice browsers that read a web page aloud.

People developing web pages have to decide which browsers to support and which standards and proprietary extensions to support.

Approach

Browser centric design vs Standards centric design

It seems to be possible to categorise web site developers in two groups:

One group seem to prefer to develop sites for only one or two specific browsers. The justification sometimes put forward for this is that IE5 is what most people use so why bother about the rest - it is time consuming to write different web pages for each browser or to write one page that detects the browser being used and adjusts dynamically to work around that browser's foibles. Sometimes this approach leads to web sites that work well in one version of a browser but which fail completely with the next version of the same browser.

Another group seem to be striving to develop web pages using standards in such a way that the maximum diversity of browsers can make sense of them. This approach appeals to me the most.

Detecting browsers

I think there are many things wrong with even trying:

It might be possible to accurately identify IE6 and Netscape 7 say, but the code may do the wrong thing in a few months time when a new version of a browser becomes available. I've seen many web pages that look for Netscape 4.5 and IE5 but which treat later versions of the same browsers as if they were unknown minority products.

Rarely will the browser detection code in all those web pages be continuously updated after they go live.

I feel too many sites do a quick test for IE and Netscape and tell the visitor "go away or download IE5" if any other browser is detected. Because of this, some browsers, such as Opera, can identify themselves as IE in order to get a chance to view the site. This leads to a kind of arms war between browser developers and web-site developers. Its ironic that even IE still claims to be "Mozilla". This dates from the days when Netscape Navigator identified itself that way and IE needed to get into Netscape only sites. This is doubly ironic now that there is a separate browser called Mozilla.

I believe that in principle its the wrong thing to do anyway. If anything you should probe for the capabilities you intend to use. You shouldn't probe to find the browser name and then make assumptions about it's capabilities.