Skip to: Navigation | Content | Sidebar | Footer


Weblog Entry

Standards

January 05, 2004

Support the standards and nothing but the standards, regardless of whether or not browsers get them right?

— or —

Support what standards are available given today’s browser support, and kludge together markup/script/CSS hacks to overcome deficiencies in implementations?

Oh sure, ‘use the best choice for your audience’. That goes without saying. But there’s an underlying philosophy that drives your decision-making process, and I know there are people who fall into each camp.

Discuss.


Reader Comments

Suzanne says:
January 05, 11h

Is there a third option of only using the standards that are supported, period?

2
Suresh JV says:
January 05, 11h

“But thereís an underlying philosophy that drives your decision-making process.”

YES. As a Designer I wanted support only for the 5+ Browsers. BUt my bosses want to support 3+ because all our customers dont have any control/option to UPgrade! ;)

What do you say?

Matt says:
January 05, 11h

At heart I’d say I’m a “standards or death” kind of guy, but in the real world you have to address your audience. However I must say that I am much more less sympathetic to browsers which appear to have been abandoned. Nothing is more frustrating.

4
Sander says:
January 05, 11h

I really think ultimately too much hacks will end the need for browser development. It?s not the responsibility of the webmaster to compensate their individual website because for faulty browsers, that?s comparable to make a special highway lane for cars who can?t get the ?round wheel thing? right and use square ones (while saying they are round).

So by introducing all these ?make up for IE flaws? techniques MS don?t have to update IE anymore because they will show customers that all modern sites still look exactly the same as on the alternative browsers.

And that should not be the long term effect of practical webdevelopment.

January 05, 11h

I typically support only the 5.x+ browsers at a minimum. For more advanced projects which tend to have a more specialized audience, only the best get in.

Roger says:
January 06, 01h

I try to go with standards only, sending browsers with little or no CSS support just the unstyled HTML, which is a lot better than locking out unsupported browsers the way we did back in the 20th century. Make sure your HTML is valid and semantic and the content should be accessible to any browser.
There are situations where I’ve resorted to using CSS hacks to get around the box model problem with IE 5.*/Win and some problems with IE 5/Mac, but I try to avoid hacks.
Using JavaScript to make IE do things that other browsers just need CSS to do seems like too much trouble.

Dris says:
January 06, 01h

Standards *are* the only tool in the toolbox. If Microsoft developed a proprietary replacement for XHTML, I sure as hell wouldn’t use it. Why? Because, in my mind, if it’s not a recommendation, it doesn’t exist. That’s just me though.

My exception to this is when a different technology fills in gaps in the W3C recommendations. For example, XUL isn’t covered by the standards, so it’s a gray area (notwithstanding the new dispute with Microsoft’s XAML). Also, I don’t feel all too guilty using -moz-insertCSS3propertyHere, because it doesn’t hurt other browsers, and it’s about to be a recommendation anyway.

I might not be correct in saying all that, but that’s my reasoning at this point in time.

When I approach a layout, I simply make it work in Mozilla and Safari. Then, I fix the things that IE/6 gets wrong using “*>#id” style selectors. At this point, the only thing IE/5 has wrong is the box model, which I remedy with the infamous box model hack.

January 06, 01h

Browser hacks are only necessary because of the dominance of one certain browser. All competitive browsers have been forced to fix their problems double quick or loose out. I doubt somehow that if the dev community started taking out all the IE hacks from their designs that MS would suddenly start fixing their browser. On the other hand, our sites would all be ignored by the largest user base out there and would fail. Shame, but its true.

January 06, 02h

kludge (external) scripts and css…yes.
kludge markup…no, if it can be avoided. maybe creatively reorganise some of the markup so deficient browsers can still display something meaningful, but i try to avoid any obfuscation only to make it look right in IE.
of course, this also depends on the “shelf life” of the pages in question. if it’s something that only needs to be up for a week or so, then i’m not too fussed about the odd bit of kludgey markup. if, however, it’s content i want to keep for a long time, i try to keep it as clean as possible for easy future repurposing.

dusoft says:
January 06, 02h

I would go with the standards - I check pages with IE 5.0+, Opera, Mozilla. Sometimes I also check with IE 4.0 and it mostly works OK.,

January 06, 02h

I use compliant markup and CSS with conditional comments for IE. I think standards-compliance is a healthy attitude that will make your site future-proof but we need the hacks to make standards-compliant sites work. The best that the current avant-garde of webdesign can hope to accomplish is a mindset towards standards, with minimal hacks to support older browsers.
There’s very little compliant sites which don’t use hacks, if only because of IE’s dominance. I don’t think you could create a successful http://www.csszengarden.com without the current hacks. Hacks aren’t only used for your audience, but also for your employers, who don’t care about compliance if your site isn’t attractive in 90% of their audience’s browsers. Thus, if you’re into web standards, you almost *have* to use hacks. Your first category seems appropriate for the overzealous who follow the W3 dogmatically.

Gavin says:
January 06, 02h

There is a 3rd (4th now?) option which I prefer to use.

Standards are awesome, and in the perfect world you could design a site with them in mind, and it would work. ;)

But since they don’t in the majority of browsers used by the population, if you can implement a hack (the box model for example) which will better your site in a fair percentace of browsers it’s a good idea.

… at my current place of work, the boss, client and the other co-workers couldn’t give two hoots if the site was designed in CSS of slapped together in dreamweaver, as long as it works. I manage to occasionally slip in the odd table-less layout still :D (just have to be weary of not spending 5 times the ammount of time building it to perfection when you have a deadline, unlike blogs and the such which are more free to do what you like).

January 06, 02h

I’m from the camp that says you should have a good grasp on standards and know how far they can take you. It should be a tool in your toolbox. Period.

The standards are a wondeful tool, but they aren’t the only tool. People often lose sight of this (although temporarily loss of sight while you are getting the hang of standards is good for bringing you to a proper state of affairs).

A bit of my philosophy (at least at one point) can be seen here: http://www.yayhooray.com/thread/7552.html

Sian says:
January 06, 03h

Websites accessible by your viewers are a must so I side with “Support what standards are available given todayís browser support, and kludge together markup/script/CSS hacks to overcome deficiencies in implementations?”. I don’t like it but thats the way it goes.

I dream of semantics, validation, standards, a much easier way of life where you’re not puzzling over which bit of code you’ll need to cobble because you know it will all work as it should anyway.

January 06, 04h

Before I meet any client…

I try to steer clear of CSS ‘hacks’ altogether – they’re messy, but worse still, they put a use-by-date on the style sheet. Sure, the ‘tantek hack’ (for example) has been tested to death *today* on *todays* browsers and can produce reliable results, but we’re left with the unknown of how tomorrow’s browsers will react to it.

How many style sheets will you all have to edit tomorrow when a popular future browser falls over at the sight of such a hack?

It reaks of the browser war era where 100’s of lines of Javascript (like the DHTML Lib) was edited and bloated even further with each new release of IE or NN.

I hide CSS2 from Netscape et al with @import, and I aim for graceful degradation and accessibility in everything else.

As for (X)HTML, I make sure it all validates, but I do conceed that content and presentation can’t be 100% separate – there has to be *some* extra divs and spans to act as hooks for styling (as per Zen Garden).

After I meet the client…

Some things require more me to bend… I won’t back down from validation and accessibility, but I will conceed to adding a few lines of JS or extra divs to make sure their drop-down menus (a recent example) work well in the current popular corporate browser, IE5/6. A good example can be found here: http://alistapart.com/articles/dropdowns/

Jeremy says:
January 06, 04h

I would have to say that you have to design for what works, since a website is dependant on it’s audience. For example, for my latest design, I had to make certain concessions in my code to make sure it worked in the horrible Internet Explorer, since alot of my visitors use that.

Yes, standards should work in everything. Yes, standards is great. Yes, I love standards. But when it comes down to it, the visitor must see the site. Especially when that visitor uses a program that a large majority of web crawlers use.

Thats my two cents, or maybe three.

Keith says:
January 06, 04h

A bit late to the discussion, but I have to say the only really answer I can think of for this is:

It really depends

Depends on the project, the audience, the client and lots of other things.

lolly says:
January 06, 04h

Right. I work in DreamweaverMX which, for the most part chucks out xhtml [transitional] compliant code. So, for the most part I comply with the standards. When it fscks it up I dive in and fix it myself. I can’t trust it to do my CSS, so I use TopStyle to do that (and very good it is too).

There is only onle place that I don’t tend to comply, and that is when it comes to placing bits of Flash content on the page. Its insert>media>flash every time. The code might be ugly but it works across the board, which is more than can be said for the compliant version (http://alistapart.com/articles/flashsatay/)…

19
Anonymous says:
January 06, 04h

Sigh … I use standards when possible.

Does anyone agree that this topic is just plain worn out? I have been designing web sites since 1998 and have dealt with most standards/browser related issues in one form or another. I find myself simply tiring of the whole topic and I don’t think I am alone here. It seems like even a lot of the big names in the business are a lot less rabid about standards these days and have moved along to more pertinent and exciting issues. Why some of them even (gasp) say tables are okay now … as long as they are used appropriately. Appropriately? Two years ago, these same people treated tables as scourges to be wiped from the planet.

Could it be that everyone is just kind of tired of the endless blathering about standards? Even after all this time, how many web sites out there are now standards compliant? Mayby 1/100 of 1%? Less? It just doesn’t seem to be catching on, does it? Even the HUGE corporate sites just don’t care. Look at their code! It sucks! But who really cares?

Standards are still good for some things though. They allow those who have made careers preaching about it to get their books published.

Or maybe its because I haven’t had my morning coffee yet …

MJH says:
January 06, 05h

In no particular order, a couple of points…

1. I recently sent a rant over to WASP (poor Steve Champeon, I apologize for my short temper) on this very topic because I’m frustrated by the amount of questions dealing with browsers and standards on the css-d mailing list. There are numerous emails each day complaining about differences in browsers, hacks, etc…

2. I happen to think that you should design towards standards, not towards a browser. I think the browsers should be the ones adjusting to us, not vice-versa. We’re playing by the rules; so should they. I’m pretty sure they have the capacity to do so. How is it so impossible for each browser to handle things the exact same way?

3. I don’t understand the concept off a “CSS hack.” Doesn’t it go the way of the tabled-design, and take away from the overall design and concept of a web-standard??? Or am I just crazy? Doesn’t it violate this: http://www.webstandards.org/about/

January 06, 05h

Anonymous wrote: “Does anyone agree that this topic is just plain worn out?”

Not at all. I do however, think that all too often the core audience that has heard the message over and over again are already “believers”, so the response is “Yes, yes, we *get* it”.

Anonymous - you asked what percentage of sites are standards based and suggested perhaps 1%. That says to me that the topic isn’t old or worn out, despite the fact that many of us have been dealing with it for *years*. We’ve only just begun…

Anonymous also said: “Why some of them even (gasp) say tables are okay now … as long as they are used appropriately. Appropriately? Two years ago, these same people treated tables as scourges to be wiped from the planet.”

Many people misinterpreted statements about using tables for layout during the past 2 to 3 years as sweeping generalizations that “all tables are bad”.

Tables for tabular data are as ok now as they were back then.

The problem is, we (collectively) see questions on lists and in various fora: “If we aren’t supposed to use tables, how can I use CSS to replicate a table type layout?”

Questions like that are evidence that people have missed the point, and we all have to provide further clarifications and disclaimers so that they understand that tables for layout is not the preferred method/best practice for site design/development and that tables for tabular data are required to provide the semantic meaning required to make sense of the data.

22
Anonymous says:
January 06, 06h


To Derek Featherstone:

I believe my guesstimate was one hundredth of 1%, not 1%. Even this, in my opinion, is being generous. After years of talk, this isn’t much of a start.

At any rate, don’t get me wrong, I design using standards and will continue to do so. I’ve invested too much time and effort to the cause to go backwards now.

Thanks!

MikeyC says:
January 06, 06h

“Support the standards and nothing but the standards, regardless of whether or not browsers get them right?”

On my personal site (source of 0% of my income): absolutely.

On my employer’s site (source of 100% of my income): absolutely NOT.

As an employee I can present the benefits of using standards to my employer, but using standards just for standards sake (eg: “We MUST validate damnit! Revenue is secondary!”) really ain’t gonna fly.

I’m a purist at heart, but in the real-world I’m a pragmatist.

Geof says:
January 06, 06h

I think that, until standards-compliance gets better, you have to keep on hacking things up a little bit. However, any hacks should be well-documented, since better compliance should allow you to remove the hacks as time goes by … assuming, that is, that one leaves the design alone long enough for better adoption of standards to make the hacks obsolete.

I think that’s one other thing to consider: design churn.

Tim Lucas says:
January 06, 07h

MJH said “I happen to think that you should design towards standards, not towards a browser.”

If creating a webpage that adheres to a few hundred pages of documentation was as time-efficient as designing towards the browser(s) then I’m sure more people would [design to the standard]…

Dris says:
January 06, 07h

Truthfully, getting sites to validate isn’t as hard as people seem to make it out to be. There are very few things that can be tougher than following simple syntax (programmers *must* have 100%-“validating” code, or it won’t even compile), and for those few things we have workarounds. Unless you’re trying to support ancient browsers (Mosaic?), at least validating shouldn’t be a problem.

Now, if you have to make a few concessions to table layouts here and there, that’s understandable considering deadlines and such.

Of course, even a site that starts out validating will likely decay over time. I’m sure I’ve left a stray ampersand on my blog many times. However, sites like ESPN and Wired show that even big commercial sites can at least validate from the beginning.

Also understandable is when reworking a site from its existing state. It can be a huge pain to make it validate, with all the cleaning up to do. Throw in archived content, and you’ve got a task that just wouldn’t be cost-effective.

January 06, 07h

Iíve blogged a bit about this at http://www.interactiveleigh.co.uk/blog/archives/000012.shtml

Itís a big issue thatís plaguing todays standards conscious web designers. Just how far can you go before you start being unrealistic?

MJH says:
January 06, 07h

Tim - good point, so - to all - why dont we have a single browser that follows the documentation to the letter… we have ones that are close, but close isn’t good enough.

I read in the summer that The WASP was moving on to phase 2 - I can’t find the article on that though. But the gist of it was ‘web standards is well on its way, and we’ve have accomplished our original mission statement.’ I’ll allow that the WASP has made people AWARE of web standards, but the standards are not being enforced properly.

I agree with Anonymous - its really not even a good start.

Mark says:
January 06, 07h

Why don’t you just throw an apple into the ballroom that says “to the fairest”?

Dave S. says:
January 06, 08h

Wow, lots of good discussion overnight. Thanks guys.

A lot of factors drive development on the web today, and whether you’re okay with admitting it or not, economics is still a very big one.

As the Little Orange Book states, standards are a continuum. You can hop on at any point. I’d been coding valid transitional XHTML for a full year before I realized you could actually build layouts without tables and bgcolors and center elements. It was a big discovery for me, and it may not be as far in the past as some would assume. It took even more time to figure out the various techniques to replicate columnar layouts and such. Should I not have been building CSS-based layouts until I had learned them all? Of course not.

Keeping both in mind, the simple fact is that if you’re developing layouts of any complexity, for a client who expects a large degree of cross-browser similarity, you eventually have to face the fact that some browsers can’t keep up as well as others. Since these browsers are still used by 90+% of the population, it’s foolish to assume that they’ll change to adapt to your code, and in that the attitude is benefitting your users.

Still, there are things that can be done to improve your situation substantially. Good points made above were using just the standards that do work consistently if you can, and keeping the hacks out of the markup and in the CSS. While some select pieces of CSS can affect accessibility, your markup is generally where the main benefits of using standard code are derived from. The CSS is the presentational layer, but the structure of the markup is what gets you your heightened accessibility, search engine optimization, and so forth.

Which is why articles like this are frustrating - http://www.digital-web.com/columns/keepitsimple/keepitsimple_2003-11.shtml - because the CSS is exactly where the hacks SHOULD go. Let’s acknowledge something: browser support can sometimes suck. Even Safari and Opera and Mozilla don’t get everything right, all the time. It’s not so much that they get things wrong; it’s that the wrong-ness is inconsistent between them. How can you overcome this? Sometimes, re-organizing your CSS and finding a new way to build it is practical. A lot of the time it’s not. CSS hacking can be useful, and to think that a hack that hides CSS from Opera 7 or displays it only in Opera 7 isn’t useful is to take a high road that doesn’t prove practical in most real-world situations.

Hacks may cause problems in future browsers? Let’s cross that bridge when we come to it. If and when this does happen, here’s why CSS hacks are better than markup hacks: you don’t have to change the structure of the site to fix your code.

As in all things in life, moderation is the key. When you build sites for other people, you can explain the issues all you want, but sometimes you just have to shut up and do what works. I’ll take CSS hacks as the least harmful way to accomplish that.

ppk says:
January 06, 08h

Dave said:

“Which is why articles like this are frustrating - - because the CSS is exactly where the hacks SHOULD go. “

Hmmm, I never thought of it that way. I don’t really agree, I think, but you make an interesting point.

“CSS hacking can be useful, and to think that a hack that hides CSS from Opera 7 or displays it only in Opera 7 isnít useful is to take a high road that doesnít prove practical in most real-world situations.”

Actually, in my specific situation avoiding CSS hacks works fine, believe it or not, also in commercial projects. I’ve never seen the slightest reason to use any hacks, except for @import, in my commercial sites. Granted, the design is not bleeding edge so it’s easier to avoid hacks, but in my day to day coding I found that I just don’t need hacks at all (again, except for @import).

“Hacks may cause problems in future browsers? Letís cross that bridge when we come to it.”

One of the theoretical advantages of the standards is that they (are supposed to) continue working in future browsers.

As most of you will know I’ve never been renowned for my strict support of the standards. I absolutely love CSS, and I use it extensively, but I rarely validate my HTML or style sheets. I’ll talk about my reasons another time.

The point I’m trying to make is that it sometimes seems as if standards promotors are talking with forked tongue. “Thou shalt follow the standards to the letter” on the one hand, but as soon as thou encounterest a browser bug CSS hacks are okay to use.

Similarly, the standards are supposed to ensure continuous working of your site in all future browsers, but CSS hacks, which empathically do NOT guarantee continuous operation, are okay to use even when you purportedly follow the standards.

Maybe it’s one bunch of people being pure about it, and other people being pragmatic about it, I don’t know, but I feel standards promotors regularly switch arguments, as it suits them.

In any case, there are three distinct parts of “standards compliancy” that should never be confused:

1) The general idea that there really ought to be web standards and that everyone, browser vendors and web developers, should follow them. I fully support this general idea, and modern CSS and DOM go a long way towards usable standards.

2) The actual standards as they exist today. I feel all XHTML standards are too restrictive and that there are some mistakes in the CSS standards (lack of support for vertical alignment, box model, etc).

3) The general feeling that web standards are the cure for all evils of the world. This especially goes for the “forward compatibility” argument, which I personally think is a religious rather than a technical argument. Standards to ensure eternal salvation? Nonsense.

The habitual mixing up of these three distinct points is what causes most of the confusion in the current standards debate.

January 06, 08h

Q: Why bother?
A: You are here, posting your comment, aren’t you?

Q: What the future (browsers) will bring?
A: I’ll think about it then.

Q: You have to use hacks if you want standard compliant web site that works.
A: Well, for me it’s still less complicated to write ‘w\idth:XXpx”, than lot of tables.

Yvonne says:
January 06, 09h

I tend to markup for standards, then hack the CSS as needed. The degree of hacks used depends on the type of site.

For a personal site, I’m more likely to let non-compliant browsers display as they will, fixing only the bugs that would affect a site’s functionality.

For a client site, it’s often hack city, particularly if I have to convert a print designer’s work (I’m a markup jockey, not a principal, so I usually don’t have much say). So far, I’ve managed to keep the hacks within CSS. I generally avoid @import, and just use every hack in the book in the main screen CSS file.

That’s this week - ask me next week and I may return a different answer.

34
David House says:
January 06, 09h

In my upcoming website, I plan on the former. (as soon as I can verify that works in Safari), prologue, application/xhtml+xml, strict doctype, all the bells and whistles.

Why? IE sucks, and people need to know it. The standards have been around a long time, and there’s absolutely no reason why IE can’t comply with them. I see no reason why not to alienate what will be a small fraction of my target audience based on their ignorance of standards. Call me arrogant, but that’s my view.

35
David House says:
January 06, 09h

Let me try that first paragraph again:

In my upcoming website, I plan on the former. <?xml-stylesheet?> (as soon as I can verify that works in Safari), <?xml?> prologue, application/xhtml+xml, strict doctype, all the bells and whistles.

Tony C. says:
January 06, 09h

What frustrates me most about this issue is most user’s absolute belief there is only *one* browser. I run across this every day at work, where I try to encourage people to at least try more modern browsers. This has nothing to do with an anti-MS stance. In fact, before Mozilla was a reality, and Opera wasn’t ready for prime-time, I went through the same thing trying to get people to ditch NS4 and give IE5.2(Mac) or 5.5/6(Win) a try.

People will drop hundreds of dollars every couple years to get the latest Office suite, because they “have to have the latest version”, but will hold on to a 4+ year old browser till their dying day, even when newer browsers are *free*.

If every designer in the world stopped doing any hacks to get their sites working niceley in IE, would people get the picture and upgrade? Doubtful. I may be a pessimist, but i think (at least at first), they will simply stop visitng the “broken” sites. Poorly designed cludges that work well in ver 4/5 browsers will win out, at least in the short term.

On a more optimistic note, I think if we keep plugging away, people will come around (including most browser makers.) The Web Standards Project succeeded in weening people off of ver. 4 browsers, thanks in part to those annoying little nags about how the browser they’re using isn’t standards based…Sure, attrition had a big part in that as well, but a great many people switched on their own, not because they bought a new computer that didn’t have NS4 installed…

January 06, 10h

Sure, I can afford the time to make most of my own website validate ( http://www.penmachine.com ) but commercial, deadline-based work is a different story.

It depends, also, on how much control you want over the final presentation. You have to accept that a completely validated XHTML/CSS design will _not_ look the same across browsers, and could be radically different in 3/4-era desktop clients, unusual browsers, and on handhelds or in Lynx. (Well, that’s true of non-validated designs too, really.)

But you can rejoice in that, and concentrate on your written content. Then again, I’m a writer, not a graphic designer, and I’m not quite as concerned about the appearance of my pages as others would—and in many cases, must—be.

January 06, 10h

I am all-for the extensive use of web standards. Anyone who’s talked to me when I’m in geek-mode, or who has read anything I’ve written on the subject knows this.

I would prefer to serve my XHTML pages as application/xhtml+xml, use XHTML 1.0 Strict or 1.1, and have all of my CSS2 work without a hitch. Some people only design for the “trinity” (which I think is a bit of a misnomer – “triumvirate” would be better), being Moz, Opera, and Safari. These kinds of things are fine for them.

Unfortunately, the real world – especially the world of business – requires sites to work as well on IE/Win as any other browser/platform. Standards are fantastic, but standards purists are missing something. We couldn’t move on from using Tables for layout until CSS was good enough to give us an alternative. Now CSS gives us that (better) alternative, but it’s not flawless – neither the support, nor the spec.

Unfortunately, Microsoft is *the* driving force on the web. Not “driving” as in progressing, but “driving” as in “this is how it’s going to remain”. I could beat Bill Gates and Steve Ballimer senseless for not implementing :hover on all elements, or :focus on anything. It makes my job as a web designer that much more difficult.

The problem lies with this: Most web designers have a passion for the web, and what it all can be. We want it to move beyond the computer screen and into everything that it possibly can. I can now syndicate another website’s content on my site thanks to the power of the web. Microsoft, on the other hand, just wants control. They want power. They want money. They don’t care about the potential of the web. All of that “Your Potential. Our Passion” is crap. Hey Microsoft! I wanna be a web designer when I grow up! Where’s your passion now?!

We, as web designers, often say things along the lines of “IE has 90% of the market. We just have to design for them. Oh well.” I disagree with this. On the other hand, standards purists chanting “everyone must code perfectly right now” are wrong as well. We need to *transition* from the older methods to the newer methods, and that requires us to design standards-compliant sites using necessary (not gratuitous) CSS hacks to make it work.

IE5 was released in 1999. IE 5.5 was released in 2000. IE6 was released in 2001. It’s now 2004. This spring will be three years since the last IE/Win update. Let’s start putting up “your browser is outdated, please upgrade” messages for IE5 now, like we do for NS4. This summer, let’s start displaying it for IE 5.5, and hopefully by this time next year a significant majority of people will *at least* be using IE6 – which will remove the need for the box model hack at the very least, maybe even more.

Promote the use of Moz, Opera, or Safari in these messages, and the Standards-Compliant market share is sure to increase. The more of us who do it, the better the chances. It might not make Microsoft update their browser, but it might make them think about it if they lose enough market share.

There’s a fledgling site that’s being put together by some people I collaborate with called DASDUA (Developers Against Standards Deficient User Agents) at http://www.dasdua.org . I don’t particularly like the name, but the premise is good. They’re working to create a website for end-users to teach them about standards and why browsers like Moz, Opera, Safari, Konqueror, and OmniWeb are better. They’re looking for people to handle different things, as they’re still growing and building. It might very well be a worthwhile cause if they can get enough support…

January 06, 11h

Ryan made a good point here. The same idea came to me when I saw “This is Cereal” submition (http://www.csszengarden.com/?cssfile=057%2F057%2Ecss). Perhaps we should design as best as possible with what IE supports, but notify non-aggressivly that there is so much better experience of the particular site (some kind of “bonus”). There is many sites that still have footer with “Optimized for…” statement…

40
Matthew says:
January 06, 12h

Personally, I like Cheah Chu Yeow’s solution to IE being a sucky browser. Take a look at his site in IE, then take a look in anything else: http://blog.codefront.net/

If we FORCE Internet explorer to REALLY suck, then people will switch. The problem lies with commercial work, however. Clients don’t like it when their site doesn’t work for their customers…

David says:
January 06, 12h

I’m with Suzanne on this for the most part. Use the standards that are supported as much as possible. You may still have to salt lightly with hacks – particularly for some of the bad box models. But there’s lots of great design that can be done only with what’s supported.

January 06, 12h

I design with standards in mind, and then hack my code up to take into account the deficiencies of the various browsers.

Not the most elegant solution, but it works.

January 06, 12h

I have not seen any news on this yet–but I wonder if MS is making any engine improvements to IE6 with either XP Service Pack 2? I think I’m going to try to find out.

January 06, 12h

I always code to standards, and then make reasonable fixes for less capable browsers - but never any that break standard compliant browsers.

45
Anonymous says:
January 06, 12h

Your css don’t validate!! What a mess…

Tony C. says:
January 06, 12h

Derek K. Miller writes:

“You have to accept that a completely validated XHTML/CSS design will _not_ look the same across browsers, and could be radically different in 3/4-era desktop clients, unusual browsers, and on handhelds or in Lynx. (Well, thatís true of non-validated designs too, really.)”

One of the great discoveries when I finally convinced work to switch our Intranet to a table-less CSS design: it started working on PDA’s (and cellphones, but that was never an issue.) The PDA issue had come up frequently, however, since the entire IT deparment staff were given PDA’s…yet couldn’t access our own Intranet to get to some resources. Once we switched to a table-less design, it degraded perfectly for PDA’s, no WML required! The same goes for Lynx…I fired it up, just to check, and it’s *much* more useable now than before.

csant says:
January 07, 02h

there is one more issue that seems not to have been addressed: we keep talking about standards refering to markup/css/script. on top of that i think we shouldn’t forget that, once we have decided for xhtml, comes the issue of *how* to serve it, i.e. MIME types… SHOULD, SHOULD NOT, MAY…aamof one browser absolutely refuses to cooperate on that. so we need to be looking for a workaround there as well if we do not opt for content denial to (alas) that one big market share… it is a constant compromise.

January 07, 03h

The only standard I respect is serving the best my clients

Clients wants fast, efficient, usable web sites that fulfils the visitorsí need. They donít mind about standards. Table or CSS, XHTML or HTML has no sense for your client.

Make your decision as a designer, and stick to it. I choose to use CSS because it:
1.make pages load faster
2.lower hosting costs
3.get better search engine results
and eventually
4.makes redesigns more efficient and less expensive
5.help maintain visual consistency throughout your sites
6.makes web sites more accessible to all viewers and user agents
7.and give a competitive edge as more of the world moves to using Web standards.

Some other may decide to use table because they are 100% web browsersí compliant. If they feel comfortable to sell & design with this message, good for them. I donít.

As a designer I mind about standards if they can improve my own productivity and create added value for my clients. As a commercial I try not to bother my clients with them. My duty is to satisfy 100% of the clients needs optimising my own resources use. This concept is called profitability. I will not accept that any standard erode my small business profitability. At the same time, it is highly respectable to me that some designers do not want to dive into the CSS/XHTML realm because they think that the learning curve is not worthy at this moment. Can you afford investing at this moment Ņ? I donít, I barely can get money to live.

Back to the CSS/web browser issue:

I bet than 99% of the clients would accept a web site delivery shorter + lower cost, if you offer him a web site IE’s + Mozilla compliant (with no commitments on NN & Opera, and remaining browsers’ compliancy). The remaining 1% should be high end clients ready to pay for a complete compliancy (actually I guess those clients would have an internal webmaster that will advocate for 100% compliancy, without looking carefully at the associated costs issue).

Have a look to others fast moving technological industries, Mobile Phones for example: Worldwide there are currently 3 dominant technologies, 3 new ones fighting for the new multimedia mobile services. Do you know any of them ? Does it really matter ? Donít you think the world is big enough for different views, different standards, different solutions, and that any of them is good enough to serve a market ? Why the internet would be different ?

To me the 100% compliancy is a non sense business concept. How many companies do you know that are targeting a 100% worldwide population reach for their products. Are you planning to serve these companies ? Standard businesses use marketing techniques to segment the market, then target their objective segments and try to optimise their sales channels for each segments ? They know the 100% reach is not feasible from an economic point of view. The CSS/Web browser issue is just about the same. Decision makers will understand this issue if you speak their language (what are the options, pros and cons of each option, associated costs of each option), not the XHTML CSS W3G 508 AAA language.

I donít think using table is a bad design practice. I do think talking technology with a Top Management Decision maker is a bad commercial practice.

Alex says:
January 07, 04h

Support the standards and nothing but the standards.

Webdesigners already do enough work as it is. An upgrade isn’t that hard, come on!

Of course, my Grandma wouldn’t know a thing about upgrading. This is a problem that can easily be solved by the method Wired uses: Have a notice at the top to warn the viewer that his/her browser is old, and how to upgrade.

January 07, 09h

I’ve noticed that some people are confusing “web standards” with “table-less layouts” etc. You can make a web page that is basically a sliced-up image dumped in a table adhere perfectly to web standards, it just isn’t the wise thing to do.

I never use markup hacks, but I do use CSS hacks like @import from time to time, but I avoid anything browser-specific.

jim says:
January 08, 10h

Just be realistic I guess. Support web standards as much as you can but know when to step back a bit. For instance, I no longer use horizontal padding in any of my designs. Until all browsers/platforms play catch up I prefer this method to utilizing any hacks. I don’t think there’s anything wrong right now in adding a few extra child divs with horizontal margins in order to save a lot of time and frustration trying to negotiate the box model problems. Sometimes we overcomplicate things unnecessarily.

January 09, 02h

“The initial premise is perhaps a bit flawed, in that most CSS hacks are fully standards compliant.”

The issue is not, I thought, whether the hack is implemented using valid CSS. The issue is that it involves *two* nonstandard behaviours of the browser it targets. Nonstandard behaviour 2 is used to feed the targetted browser CSS directives to compensate for nonstandard behaviour 1.

Now, what happens when a new version of the browser in question comes out? 4 possibilities:

1) Neither behaviour is fixed.
2) Behaviour 1 is fixed, but 2 is not
3) Behaviour 2 is fixed, but 1 is not
4) Both are fixed

Case 1) and 4) are OK: the browser will either behave just as before, or it will behave (in this regard, at least) like any other standards-compliant browser.

Cases 2) and 3) are rather unpredictable. The site will most likely break in some nasty way. And, since years have gone by, you may or may not be around to fix it.

January 09, 12h

The initial premise is perhaps a bit flawed, in that most CSS hacks are fully standards compliant.

At least one commenter stated something about “avoiding CSS hacks” but “rarely validating”. This seems completely non-sensical: religiously anti-hack and yet flippantly invalid?

dan says:
January 17, 09h

I say Standards with hacks. Well hacks only when needed.

I test in ie 5+, NS5+, Mozilla 1.5.

I figure if someone isn’t using one of these or a compliant browser they know what to expect at this point. Besides the main issues I would say is Netscape 4 and with all the crap that is out there, seeing a text based site is better than a badly designed site anyday. If I’m looking for information and I go to a really “bad” site, I might not stay for too long, but if it’s just text, I will go through it.