I’ve put together 5 basic levels of IE6 support. Which do you aim for when developing a new site?
1) Perfection
Visitors using IE6 get a pixel-perfect, fully functioning site, just like visitors with a modern browser.
2) Pretty close
With some minor differences (e.g. alpha transparency, fixed elements, minor positioning glitches), IE6 visitors get a decent facsimile of the site modern browsers enjoy.
3) It works
The site ‘looks broken’ in IE6, but visitors can still perform all functions and access all data.
4) Good luck
Site hasn’t been IE6 tested at all. It will probably provide basic functionality, but if it’s not working, it’s the visitor’s problem.
5) Gulag
IE6 visitors are not served any content, and are forbidden to use the site.
I tend to aim for between 3 and 4, with a warning message at the top of the site. What do you do?
Sorry, I wasn’t impying that you or any particular people were going for broken pages. What I was trying to describe was the “type 3” mindset (which is one that I share, to a greater or lesser degree), which is “I will aim for a site that works fully and as intended in all modern browsers, and that works sufficiently in legacy browsers”. I never set out to create a site that is broken, which is the way I read your words - I set out to create a site that works, but the bad browsers break it. And sometimes there’s a limit to how much hackery and trickery I can be bothered to go into.
You seem to share the same attitude. Do your clients know this?
I don’t do paid web design work. The projects that I do are personal or pro bono.
Thanks for sharing the link, it is a prime example of what happens when you don’t set “haslayout” for IE6/7. It just so happens that was the very first thing I mentioned in my list of common things that break IE6/7.
You had way more than just one issue in IE7, it actually looked worse than IE6 to me.
That’s the thing I don’t get about IE7 - not only did it fail to fix a lot of IE6’s problems, but it actually introduced a load more new ones…
With just about 30 minutes worth of reworking your CSS I got the page to what I would consider level 2 in the Op’s scale. I only worked on that homepage though, I’m sure the others could be fixed just as easily.
Thanks for taking the time to do that. I will have a look at what you’ve done, try to understand it, and then see if I can incorporate it into the site On first glance, it looks like some of the spacing has changed in ‘good’ browsers, so I’ll have a play and see if I can get it to do exactly what I want first.
For me the answer is, “It depends.” It depends on who is using your site, what browsers they use, and the level of support that is appropriate for your audience.
In general, I think it’s smart to start with the Yahoo Graded Browser Support system (http://developer.yahoo.com/yui/articles/gbs/). You serve JavaScript and CSS to browsers on their A-list (which is constantly changing, and btw currently still includes IE6 on Windows XP). You then have a somewhat solid basis for all the browser/OS permutations you’re going to support initially.
Once you site is live, then you can track usage statistics and see what makes sense for your site. Every site will have its particular flavor of users using a variety of OS/browser combinations, and you tailor the site to that flavor.
Sorry if I appeared to come down hard on you, it really wasn’t my intention.
After reading back through my post I can see how it may look like that.
The point I wanted to impress was that the majority of IE problems are related to haslayout. When people understand how haslayout works they can cater to it while writing the code. It does not take much time at all.
That’s the thing I don’t get about IE7 - not only did it fail to fix a lot of IE6’s problems, but it actually introduced a load more new ones…
IE7 is really just a glorified version of IE6, what it did fix was #3 & 4 in my list. But it left us with some other bugs that are not seen in IE6 at the same time.
haslayout
IE broken float model
Double Margin Float Bug
Expanding Box Bug
I will have a look at what you’ve done, try to understand it, and then see if I can incorporate it into the site On first glance, it looks like some of the spacing has changed in ‘good’ browsers, so I’ll have a play and see if I can get it to do exactly what I want first.
I put a lot of comments in the CSS so hopefully that will help you see what I did. As far as the spacing, I set an even 25px between the three columns. You had the left and right columns at pixel widths with em side paddings. You would get unpredictable widths with different font-sizes on the user end.
If I increase the font-size in Firefox (ie, if I zoom in), the container grows but doesn’t truncate the content. If I do the same in IE6, it seems as though you’re employing a maximum height and the content expands beyond the limits of the container. In other words, content gets chopped. I’ve checked this in normal resolution (1440 by 900) and full screen (1920 by 1200).
One of your images, http://www.hitechy.com/images/design_hitechy_1.png, weighs in at 203kb. I’ve just checked this in Photoshop and found that it should consume no more than 79kb as a PNG; 90kb as a GIF; and 93kb as an 80% JPG. You need to learn the fine art of compression.
My website makes use of the :target CSS3 pseudo so I spent hours retroactively developing it so that IE6 users would get the most primitive but perfectly acceptable experience (where none of the cool tab swapping effects occur and everything loads in the one section). It’s by no means a perfect solution (in that some people may find it less appealing) but IE7+ users can take advantage of the JavaScript supplement that turns the IE6 support level to that of any other browser (I replicated the target pseudo progressively) and thus it works and works well. As far as I’m aware I’m the only person on the web to have successfully scripted a usable workaround for the target pseudo and it’s various complexities, I’m actually rather proud to say I’ve a CSS3 heavy site that works in IE6.
These all seem like appropriate decisions to me. Particularly the first one, if you have a site for a limited audience, it’s completely appropriate to design to your audience, not the general public.
All the other cases you cited were also textbook progressive enhancement issues. Provide the appropriate level of experience that each browser can handle, IE6 will get the level of support that it can handle, IE7 will get the level of support that it can handle, safari will get the level of support that it can handle. Good design is about giving to each their own, and progressive enhancement has been so successful because it’s the simplest most manageable way for us to implement that mandate.
[I]I’m also right with you on the javascript work arounds. Implement those wrong and you’ve actually destroyed your site rather than enhancing it - progressively or otherwise!
Though the Sitepoint book I just got (jQuery: Novice to Ninja) has some excellent tips for handling exactly these kinds of issues the “right” way and taking advantage of jQuery utility features to automate much of the hard work for you. I highly recommend it. If any of you have been tempted by the power of jQuery, IMO this is the book that shows you how to take command of that power and turn it into a real game-changer for building accessible, high end RIAs.[/I]
I couldn’t agree more. IE6 is not capable of handling the high end CSS needed to make first class sites these days. It’s just not, never did, never was meant to, that’s the perfect scenario where you make sure the browser supports “as much as it can take” and that nothing higher up breaks any of that. This is progressive enhancement territory, and IE6 isn’t going to take as much enhancement as FF, O, IE8, and other modern browsers.
On the flip side, I really don’t get why this is such a big conversation. I write all my stylesheets strictly in good CSS aimed explicitly at modern browsers. I do not introduce hacks to make IE6 work, I don’t even look at IE6 while I’m putting the design together, actually, because I don’t want to screw good browsers under the temptation to make 6 look better, so I don’t even look at it. Once every “good” browser is working, I look at IE6 and crack open a stylesheet aimed at IE6 and below by conditional comments. I look at all the different browser behaviors from zooming to resizing, etc., etc. and identify any trouble spots. I then selectively deactivate or modify those elements in the IE6 specific stylesheet. This requires…
Validating your webpage at W3C (which always a good idea).
Validating your stylesheets at W3C/Jigsaw (which always a good idea).
Doing a quick check in IE6 (and modifying the conditional stylesheet, I think this step took me a grand total of about 30 minutes for my last site)
Now I’m not the fastest designer on this forum, or the biggest, or the most creative, the most innovative, or the most experienced. But even I think that’s pretty freakin’ easy, and your main stylesheet is “pristine” and built exclusively for modern browsers along W3C preferred specs.
That’s all the “investment” we’re talking about. Given the number of visitors you can garner, there are few opportunities to make so much difference to your website for so little effort.
I guess it depends on your perspective and whether you’re developing websites in the best interests of your clients or according to how much effort you feel like putting into your work. To be clear, the last thing I want to do is write the IE6 stylesheet that every site I’ve built for the last four or five year has. In fact, it is usually the last thing I do but as Alex mentioned, it’s not that difficult once you’ve done it for a while. To me the reason I do it comes down to numbers.
The websites I find myself building aren’t Amazon, Google, Digg or any really high volume sites. They usually serve anywhere from 2,000 to 6,000 unique visits per month according to Analytics and AWStats. That said, I can quickly pull up an IE6 report with either stat program that shows in some cases up to 10% of the visits were using IE6. The visits in IE6 for most of the volume sites is about 200 - 250 but a report on one site which belongs to a hotel indicates that over 500 visits were with IE6. I don’t know what the rate of conversion is for them just on IE6 but I’m sure they would be happy that their potential clients see their website in it’s best light, regardless of whether they are using an outdated browser.
Ive only had to use the “fixed” position once recently but there is a good little hack that does the job with IE6 conditional includes and a bit of editing (http://www.howtocreate.co.uk/fixedPosition.html).
Also I’ve been using http://spoon.net/browsers/ for my testing, bit of a resource hog but a lot easier than having ie6, 7 8 installed on my computer at the same time!
So yeh I say for the time being go for 1 and worst case aim for 2. Generally the clients I work for would not be happy with anything less than 2!
Well it must pay off because I’ve seen your site many times on those inspiration galleries
Actually, most of them probably think “What’s IE6, what’s a web browser and what does my audience have to-do with a design?” if the stereotypes are accurate. I think it’s a bit of a shame you turn down customers on the basis of needing IE6 support (seems a bit browserist - if that’s a word!), it doesn’t take much to get IE6 having a simple basic experience, a conditional stylesheet to correct the wonky bits and pieces is very easy actually. Those have eliminated the need for most hacks and thereby, much of the complexity if you’re trying to feel your way through the experience. Though I must wonder how you know the target audience don’t require IE6 support… even with good statistics software there’s a very high margin for error.
1-2 years ago, I was still fully testing in IE6. Since January, I’ve pretty much dropped it.
I will check to make sure nothing major is broken in IE6, but I won’t spend much time with minor stuff to get a pixel perfect layout. Considering certain giants on the web have dropped IE6 in addition to the low percentage of users actively using it, I don’t find it worth the time or aggravation… But, if a client’s base of users does actively use it, then yes, I will test thoroughly with IE6. Within the last couple years, that hasn’t happened, though.
Andrew’s list is pretty much what I actively support these days. I just added safari and chrome within the last few months, so I’m still a little new to them (though they haven’t given me any major headaches yet).
Granted, I will still do a spot check with older and obscure browsers using browsershots.org to check for major issues.
Netscape 4.xx. Those were the days. There are too many issues involved in supporting NS 4 & IE 4 to even go into it. What a nightmare and then MS released IE5 or 5.5 (I don’t recall) and it was like magic… Really easy to make pages with and it would make up for any mistakes we introduced into the code. We were hooked. I remember when I was playing with Mozilla & Opera browsers, trying to sort through making consistent html/css, a buddy of mine saying he didn’t care about any browser except for IE5.5 or IE6 because he didn’t have to write clean table code. IE would just guess at what he was doing and make it look good. It didn’t matter to him how it looked in anything else. Super nice guy but we didn’t share the same work ethic
Thanks goodness for for articles about standards compliant coding practices and IE CSS hacks on SP… That was all that saved me from slow and painful IE withdrawals.
They probably say to themselves “It’s a damn good job most of my potential customers / target audience don’t use IE6 and I don’t have to pay extra for support for that browser that is almost a decade old. Good job most of my potential customers / target audience have enough money to realise that once you get into the technology arena it’s a world of updates and upgrades that don’t come cheap and they’re willing to upgrade their operating systems for a better Web browser, amongst other features.”
No, I tell them what Web browsers I do support and test in and also ask them about their target audience and do some research to see whether I’ll need to create a site that has support for IE6 and then when I do find out that they require a site with a target audience that uses IE6 I refer them to someone else who is willing to lose some hair over it Everyones happy then.
As for the second question I’ll quote myself because you must have missed my list of supported Web browsers which are:
Thanks for the tip, I already do :tup: I think I’d have some complaints if my client’s target audience complained about their Website but they must like the Website because my clients give me some really nice feedback telling me how their own clients love the Website.
I have access to a computer running IE8 and a computer running IE7 and I have a Microsoft Windows Vista operating system. I don’t have IETester installed though and although I thank you for your recommendation I’ll have to pass. As I stated earlier I don’t support IE6 and don’t intend to in the future. I won’t go into a long post about more reasons and my argument for it because it won’t make the slightest difference and I’m not here to argue about IE6, I’m in this thread to simply state my level of support for IE6 which I have done.
On the other hand, Dr John, if you have used a Website that has been designed / developed / produced by me in the past and you were using IE6 and you had a problem with the Website then please feel free to make a complaint to the owner of the Website or myself with some details on your problem and I’ll get to the bottom of it for you promptly.
Kind regards,
Andrew Cooper
[Side Note: Posted by a dedicated and loyal Internet Explorer user and Microsoft fan-boy. I just don’t support IE6.]
Your clients must be REALLY reassured that you don’t care what THEIR new site looks like for the poor IE6 users. They probably say to themselves “I wouldn’t sell my goods / services to an IE6 user and I’m glad I found a web guy who just says Screw You to those potential customers of mine who are stupid enough to use IE6”
Do you tell your clients that you can’t be bothered checking the site in IE6? do you check it in Chrome or Safari, which have a similar market penetration?
Design a site for the visitors, not for your own computer.
You don’t have access to a computer running IE6??? So you don’t have access to a computer running Windows and don’t have IETester installed (google it - I’m too arrogant to give a link to someone who doesn’t support IE6 - jeez I’ve just told you the name of the software to use, perhaps I should delete that and let you guess it…)
Mark, I hate to sound arrogant / ignorant / cocky / jerky / whatever but I’ll have to say that if that is the case then I don’t care for a lot of people (according the the statistics of IE6 users on the Internet) although it just so happens to be that the majority of visitors to the Websites I design / develop / produce use IE7 or IE8 rather than IE6.
For me it comes down to some core principles and concepts surrounding the technology arena, but that post is for another thread!
I also agree with you milenko1054 Most of the times when I have designed / developeds / produced a Website and tested in the five major browsers (IE being IE8) I will usually test IE7 last and it will only usually be a tiny error that I can either live with or just edit one number say from margin-top: 20px; to margin-top: 10px; which is usually the case. I said I don’t test in IE6 but as I also said, I can’t see any major out-of-whack errors and if there is then I couldn’t care less Helps me keep my hair on my head!
On first glance, it looks like some of the spacing has changed in ‘good’ browsers, so I’ll have a play and see if I can get it to do exactly what I want first.
