5 Most Common Web Standards Misconceptions

Delicious | Digg | Stumble | Reddit | Subscribe to RSS Feed

When starting to get interested in web standards, you hear a lot of buzzwords, a lot of theories about what web standards are... Here are some most common misconceptions.

1. Web Standards = Validation

Markup and CSS validation has an important role. It can be of great assistance while you're learning, because W3C validator not only points out the errors in your code, you can also learn how to fix them.
However, passing validation alone doesn't mean that you're page is built according to web standards.
I once compared validation and spell checker. If you run this sentence through spell checker

Dog fox brown lazy over jump.

it will find no errors. But does that sentence make sense? No. Similar thing happen to your document when you're not using proper html elements for structuring your markup and not paying attention the the logical flow.
I can't stress enough the importance of semantic markup.

2. Web Standards = CSS

CSS is a style sheet. It is a (powerful) tool that we use to make things look good. And yes, it is very important part of web standards.
But, take a second and look at any table-based markup nightmare of a site and you'll see that they all use CSS in one way or another. Does that make those site web standards compliant?
Technology doesn't make web standards. Even CSS doesn't. It's how you use it.

3. Web Standards = No Tables

When switching over from table based layouts to css based layouts, many developers refuse to use tables in any case. "Tables are evil" they say.
To me, none of the html elements are evil. They were all invented by very smart people and each of those elements has its purpose. HTML tables, for instance, are the best choice if you have to display tabular data, such as calendar...

4. Web Standards = Divs

Overusing divs is something that often happens to people switching from table-based coders to css rookies. I went through that phase myself. What happens is, you simply replace tables with divs and that's it. You end up with the similar complex and unreadable markup as you had before, plus your css file grows several times.
It's not just the container elements that you have to change, it's the entire approach. It's quite a path to walk, but it's the good path :)

5. Web Standards = No IE

The idea of web standards is delivering cross platform and cross browser solutions. In IE's case we need to deliver web standards to non standards compliant browser. But it's our job.
Do yourself a favor and don't ignore IE debugging when developing your site because your client use it for sure. Sooner or later you will have to do it.

About the author:

cssglobe's image Designer, developer and a passionate standardista with huge experience in all types of front-end work. I've been featured on numerous css galleries including famous Css Zen Garden official list. Drop me a line if you have work proposition or if you'd simply like to say hello.
To stay in touch, follow me on Twitter.

Enjoyed the article?

Then don't miss our next one! Subscribe to our RSS feed or share it with your friends.

Delicious | Digg | Stumble | Reddit | Subscribe to RSS Feed

Comments

Mario Andrade on 22 Mar, 2008 wrote:

There's a big problem with IE, many time i had to develop practically 2 different style sheets to work correctly in IE6 and Firefox, than IE7 came out, the same site would have problems displayin correctly in IE7, the easiest solution was to create a new stylesheet for IE7 specifically.

Now IE8 is coming and I already tested and at least 4 of the websites I designed don't display correctly.
I put the code on the w3c validator and the only invalid code was flash+dreamweaver code.

Altought it's our job and every website should work correctly on every browser, maybe we should set a limit.

All my clients had IE6 in 2008, i advised them to update. Many problems where solved there.
Bedrich on 22 Mar, 2008 wrote:

@Mario
If you are feeding each browser a different stylesheet, you may want to revise your CSS. And when you state that your stylesheets are "almost" valid, you are falling under one of the misconceptions this article talks about. Valid code does not equal web standards or proper coding methods. Lastly, your clients upgrading their versions of IE does not fix many of the problems. There is still a big chunk of users browsing the web with IE6 (about 30%).
Steve Firth on 22 Mar, 2008 wrote:

Agreed, for common and garden layouts you shouldn't need hacks and certainly not dif styles for dif browsers ... aside from the v occasional * html to fix a width when I'm feeling lazy.

As regards "div soups", I was updating a form page someone else had made. There should have been 7 divs, 10 at the v most. Actual div count 207, I died a little inside.
Matt Basta on 22 Mar, 2008 wrote:

Mario: I disagree with you entirely. Most problems in a well-designed website can be fixed in IE by simply making two CSS statements on the particular object. Having a seperate stylesheet is not only wasteful, but time consuming and confusing for anyone updating the site in the future.

IE8 is beta software. There is still a lot of time before the final product is released.

Where I work, not supporting IE6 will get you fired. I agree that sometimes the functionality is just not what you would want, so graceful degradation is sometimes necessary. It is no excuse to build a website that simply doesn't work properly in IE6, though.

If every one of your clients has IE6, then it isn't your place to have them update. It shouldn't matter how it looks on their end, it should matter how it looks on your clients' clients' ends--something neither you nor your clients can control. No offense, but building a site with this paradigm is simply lazy, cheap, and arrogant.
Ofer Chermesh on 22 Mar, 2008 wrote:

I established with 3 friends an Internet service company named Ghotit (www.Ghotit.com) that helps people with dyslexia move to the main stream which is the general population that don't have any problems to read and write.

After investing more than a year on developing on our solution we started to receive very good inputs please try Ghotit and I hope you will love it
Harry Roberts on 22 Mar, 2008 wrote:

Very good write up. And very true. Nice one.

//Harry
Fabián on 22 Mar, 2008 wrote:

There's a pretty good alternative to write alternate style sheets: unobstrusive javascript.
This way you can have beautiful html and css and then fix the IE bugs with say.... JQuery!
suraj naik on 23 Mar, 2008 wrote:

the article is good enough, but as others are fighting on the subject IE, i would like to tell them that..if you write a clean markup & if your css is that much strong..IE render the same way as others can..some time better than that..
cssglobe on 23 Mar, 2008 wrote:

IE is something we have to live with and deal with, no doubt about that. IMHO, mistaking validation for web standards is a greater misconception.
Dana Lee Ling on 24 Mar, 2008 wrote:

Excellent misconceptions list. I do try to implement both the letter and spirit of the standards as best as I can, though I fall short often. The first misconception is at the heart of my current conundrum. I happily tossed out the "i" element in favor of "em" back in HTML 4.01 Strict's infancy. Now HTML5 has me rethinking this: botanical names are italicized. Always. This is not a stylistic choice but a semantic rule: Latin is italicized to denote the linguistic difference. There is meaning in this difference, and the rule is italic, not "emphasis." My point is that while validation does not guarantee standards compliance, defining "semantic markup" is going to be a difficult task at best. One person's semantics will be another person's presentation.
Fahed on 24 Mar, 2008 wrote:

I was hoping to get a clear definition at the end. Any chance?
Andrew Assarattanakul on 24 Mar, 2008 wrote:

Joel put it well in his article Martian Headsets (http://www.joelonsoftware.com/items/2008/03/17.html) about the major problem we have with what we call Web Standards. It isn't really a standard because browser makers do not have to conform to it.

W3C can only recommends what browser makers should do and how it should behave. But the problem is that they aren't straight forward and clear on how it should be in certain cases and so each browser maker ended up doing their own thing and interpreted them differently.

@Mario: For IE8 submit the rendering bugs to their development team. It is not done and probably isn't anywhere close to the way they want it to be in the end. There are far too many rendering problems and specs missing in their new rendering engine to be considered finished.
Aaron Hildebrandt on 24 Mar, 2008 wrote:

I think people also have misconceptions about how broken IE6 is. Yes, it has a lot of strange quirks, and I'll be jumping for joy when we can finally stop supporting it, but it's gotten easier and easier to accommodate it as time goes on. Once you've learnt the quirks you have a good idea of what works and what doesn't. Now, I can usually code something that looks fine in every major browser in one go, completely without hacks. Granted, I stare at HTML 8 hours a day for a living, but once you've learnt what IE6 does and why, the idea of serving separate stylesheets is ludacris.

My biggest peeve in web design: people who write invalid code but claim it validates because it's inserted via JavaScript. Okay, that might not be my BIGGEST peeve, but it's up there.
Logan Capaldo on 24 Mar, 2008 wrote:

@Dana

One possible solution is <span class="Latin">
anon on 25 Mar, 2008 wrote:

I'm sorry...but IE is crap. It's true that until the industry says, "your product stinks and we refuse to support it", will that change. Until then we will be forced to duct tape our code to make IE render the way the rest of the world does. (ex. IE 5 .png transparency) I continue to use these hacks (duct tape) until that time...but deep down inside I know most coders just want to say F#$&* You IE!
Agent 001 on 25 Mar, 2008 wrote:

I like that. What you told about IE is good. Almost 70% users on Internet use IE.
Ravi Rao [MSFT] on 25 Mar, 2008 wrote:

@ Mario:

If you have rendering problems and other layout issues with IE, feel free to run them by me, and I'll take a look. My email is ravi@alumni.concordia.ca

We're actively listening to our customers, so please do not hesitate to give us feedback via the IEBlog, or the Beta-1 Newsgroup, or via the Connect Bug database, or any of the various other means.

Thanks.
Malene on 25 Mar, 2008 wrote:

I agree with the article, but I can't understand why some of you are against dif stylesheets for dif browsers? I'm still learning all the get arounds to work out IE6 css, and until I've mastered them, dif style sheets are the only way to go for me. At the same time I would lkike to ask, if somebody has some good links I could visite to learn to make only ONE css:D
Jermayn Parker on 26 Mar, 2008 wrote:

If you code using proper CSS and HTML, problems in IE will only be minimal.
Dallas Web Designer on 1 Apr, 2008 wrote:

This is great to know. I know that when I a test sites that I am creating I debugg in Firefox and IE. Going to a no table site can be tough at first but its well worth when you are done and see how fast the site loads.
Jitendra vyas on 7 Apr, 2008 wrote:

very good article. good comparison

Sorry, further comments are disabled for this post.

Hey, you are not logged in!
Apply for a membership or login here.

Want to join?

If you wish to contribute with your own articles, updates or gallery entries you can apply for a membership and even become our editor.

Apply here

Css Globe