Funny how people usually blame the non-IE browser

[DrDoc]

Is there anything we can do about this? People usually blame the non-IE browser when something doesn't look/work the way they want it to, when, in fact, IE is the browser rendering the page the wrong way. And, more often than not, if you make the page work in the non-IE browser, IE still displays it correctly.

Why do people design for IE first? I don't get it!
I view the page in IE, but I sure don't use it as the corner stone of any development.

[shasan]

It's the same reason why a lot of new houses in Vancouver, Canada are built to be Feng Shui 'compatible' (huge asian population)

There are two types of customers:

1. Those who care about Feng Shui
2. Those who don't give a darn if it's Feng Shui or not.

If you don't develop using Feng Shui, you lose out on the people who care about Feng Shui.

If you do develop using Feng Shui, you lose out on nothing.

Might as well cover all bases no? It's not that tough to write valid code.

[BjarneDM]

PCInk

Can anyone give us an example of any W3C validated code that does not work on all IE and NN and Opera browsers?

That's an easy one: http*//mozilla.mathiesen.info/

The reasons it doesn't work in IE is
1) there's an xml-declaration at the start of the document, which triggers IE6 quirks mode
2) the server sends the documents as application/xhtml+xml which is the recommended way, but IE6 only understands text/html.

Except from the use of -moz-border-radius, the css *is* valid. I dared use it because w3c has a border-radius on the drawing-board for css3

On Mac OS X
- IE5.2 wants to download the page, and won't display it at all
- Opera 6.2 doesn't apply the style-sheet

[D_Blackwell]

---Can anyone give us an example of any W3C validated code that does not work on all IE and NN and Opera browsers?---

I do it all the time. Just because it validates, doesn't mean it won't need tweaking. I validate most everything, but it's pretty easy to make a layout that validates and breaks. (Usually pretty simple to fix though.) If only I were experienced enough to code around all the pitfalls:(

[asquithea]

I write all my pages in XHTML 1.1, and send them using an application/xhtml+xml mime type to browsers that indicate that they support it. The most immediate benefit of this practice this that NO invalid XHTML can escape validation, provided I actually view the page before releasing it. Most browsers will mask errors even when the DOCTYPE requires strict interpretation if they are still receiving a text/html.

Validation alone is insufficient to ensure that a page works. This page [in-solutions.net] for example, wouldn't work even if it validated, because the author failed to understand how whitespace in elements affects their width.

This second problem affects me constantly with IE. I cannot, for example, design a site that uses a fixed CSS menu without a hack or some browser sniffing, because IE does not understand that property. Nor can I use many CSS2 selectors. The CSS might be perfectly valid, but that doesn't help me. I cannot rely on the <abbr> tag, for example, because IE doesn't support it, even though it is valid XHTML. Neither Firebird, IE nor Opera seem to suppport the longdesc attribute for images.

So now, as it always was, graceful degradation is key. I design my sites to work with the newest of everything from screen readers to the latest Firebird build, beginning with the best technology, and working down to the least capable. IE6 comes well down the list.

I do NOT test and fix for old browser versions in any significant way, however. I believe that one of the reasons M$ have gotten away with a broken browser for the last 5 years is that users cannot see the breakage. Users of my sites will see a perfect site in a fully standards-compliant browser, and a slightly less attractive (but perfectly functional) site in IE6. Screen-readers and text browsers will have no problems at all, not being able to see the CSS, which causes most of the problems. I provide a slightly customized help page for IE users that explains why the site looks slightly odd, and where they can download a copy of Firebird ;-)

So, to sum up, I can get away with slight breakage in IE and old browsers because my sites don't need to make money. Were I worried about that, I'd do everything the same, but add a patch CSS file at rendering time based on user agent.

[Shadows Papa]

Had to respond to the example of the INS page.
It may have validated but it's not very good from a "person's" viewpoint.
All the wasted space, and I had to wait for the "powered by" logo - what a waste. Like I'd care. I'm not a fan of "powered by" logos. It's simply free advertising for someone, not a sign of company strength or professionalism.
It's too large for the most common screens - 800x600 (I have horizontal scrollbar when viewing it - no one should ever have to scroll sideways at 800x600)
Bottom line, they could lose a graphic or two, clean out the white space in their code (they are using TABS to space their lines out nice and tidy) - AND on the displayed page, make the page scale a bit better for common folk, declare a doc-type, lose the old <center> and <font> tags, lose the spacer images and tables that are set a hard 800px wide...
I'm no pro, but they aren't either, and it shows.

Ha, I guess it's good to see a professional networking company do a web page the old way, too!

Shadows Papa