Tuesday, July 22nd, 2008

Pondering Support of IE6

Category: IE

If there’s a single experience that bonds all Web developers, it is the incomprehensible joy of wasting hours of our lives supporting IE6. The quasi-palindromic Dionysios Synodinos (alright, it’s pretty far from reversible, but there’s got to be a good anagram in there) over at InfoQ has put together a fun piece discussing wither IE6. From the article:

Since attaining a peak of about 95% usage share during 2002 and 2003, Internet Explorer 6 (IE6) has been rapidly losing market share. As the end of 2008 approaches, significant online services, vendors and web frameworks are dropping support for IE6. Will this year be the end of IE6 and what does this signify for Web 2.0 developers?

Dionysios draws examples and quotes from Apple, SproutCore, 37signals, Douglas Crockford, Dylan Schiemann, John Resig, and more as he searches for an answer. There’s no conclusion to the piece, but plenty of perspective. Some excerpts:

Dylan Schiemann of Dojo: I think we’re stuck with IE 6 for at least another year.

Douglas Crockford: If you are operating a mainstream site that appeals broadly to all users, then regrettably you must continue to support IE 6 until its share drops to insignificance.

Jeffrey Zeldman: If your company can afford to quit supporting IE6, now is as good a time as any to do so. Whether or not you can do so depends on your audience and business model.

If nothing else, raising the hope of dropping IE6 support before I send my kids to college makes it a welcome article indeed.

Posted by Ben Galbraith at 7:00 am
37 Comments

++++-
4.5 rating from 36 votes

37 Comments »

Comments feed TrackBack URI

Pffft.
60% of our hits come from IE6. It’s those banks and large business enterprises that don’t switch because they’ve got ActiveX controls written in IE6. Employees don’t know better, and CTOs don’t want to take the upgrade risk. Ironically, the IT staff (cept the activex control devs) all use firefox cos they know better.

Firefox should target companies with their marketing campaigns and Microsoft should strong arm enterprises to upgrade. By strong arm, I mean threaten with molotov cocktails.

Comment by CVertex — July 22, 2008

Here’s another quote for you, “dropping support for IE6 is a moral obligation either your business model can afford it or not. When dropping it you can sniff the browser string on the server and redirect to http://getfirefox.com
Thomas Hansen

Comment by polterguy — July 22, 2008

http://www.infoq.com/news/2008/07/ie6_on_its_way_out

Comment by deanedwards — July 22, 2008

http://www.infoq.com/news/2008/07/ie6_on_its_way_out

Comment by FrankThuerigen — July 22, 2008

Same second Dean ;-)

Comment by FrankThuerigen — July 22, 2008

Now if there were only a special META tag that would bring up a dialog that says:

THIS WEB PAGE REQUIRES THE FIREFOX PLUG-IN FOR IE6.

We could make the world a better place.

Comment by Diodeus — July 22, 2008

I work for a firm that caters to government officials, whose offices have very strict guidelines regarding browser compatibility. While we can encourage staffers to use IE7 on our intranet applications, dropping IE6 support completely for our public websites is financial suicide right now.

That said, I would have no problem encouraging users on any sites that I develop outside of the company to upgrade IE. The efforts of Facebook et al. will be what drives people to finally ditch that old beast.

Comment by elfpoet — July 22, 2008

I think those people that still use IE6 are those that are forced to do so (by company intranet policies / activeX ) or those that don´t feel the urge to use high-end web-apps anyway. While I still support IE6 with my lib, I wouldn´t do if I had to rewrite it now. Its good to see IE6 vanish, its faults bind way too much developer time all over the world.

Comment by FrankThuerigen — July 22, 2008

I charted IE’s version popularity over the last 9 years with data from TheCounter.com a little while ago:

http://remysharp.com/2008/03/11/when-should-we-stop-caring-about-ie6/

The hope being that if IE6 follows the same trend as IE4 and 5 did, then, by my interpretation, IE6 will be in the ‘insignificant’ figures by 2010.

Comment by remy — July 22, 2008

All good initiatives but Apple, 37signals are not the once that decide end of life for IE6, Microsoft is. Since XP SP3 shipped IE6 we can whine about it for at least 3 more years.

Comment by Jadet — July 22, 2008

Yes, IE6 is pretty bad. PNG alpha channel support, the box model, XHR…most of it is missing or terrible.

But it’s not that bad. There are now fairly standard ways to work around IE6’s shortcomings, so I don’t really mind that it might still be a concern a year or two down the road.

Though I wouldn’t mind if it just magically disappeared somehow.

Comment by MichaelThompson — July 22, 2008

Microsoft needs to step up to the plate here. Enforce the auto update to IE7 that they tried when IE7 first came out. Now that it’s stable, I’d be for it. And when I say update to IE7, I mean 7, not 8.

Comment by leachryanb — July 22, 2008

All we do is deliberately *leave off* the IE .png fixes, stylesheet tweaks, etc. etc. People will soon get bored of rubbish-looking web sites.

Comment by spyke — July 22, 2008

The answer is pretty simple. Have the site degrade properly for IE6 users, but let the A-grade browsers enjoy all of the benefits, whether it be animation, AJAX, amongst other things.

Comment by Liquidrums — July 22, 2008

Anyone who doesn’t support IE 6 should be a personal hero of every Web developer world wide. They deserve recognition.

We should find a way to reward those chosen few who have demonstrated the courage to make the world better for the rest of us.

The best way would be if a prominent Web Development / JavaScript blog could host a list of such people. Rewarding them with extra traffic for making that choice.

But what blog could do that? What blog might want to lead the charge against the bane of IE 6.

Comment by JustinMeyer — July 22, 2008

A friend a mine who used to work for the govt. tried to upgrade over a 1000 PCs to IE7. Then things started to break. He had to revert back to IE6 because they didn’t have the resources to fix the problem. Plus, most of the employees didn’t actually want to put forth the extra effort to fix the issue. Seemed to be the theme during his employ. He had great ideas, but everyone else wanted to maintain the status quo.

Anways, I’m plagued by IE6 with every project. Can’t wait till it falls by the wayside.

Comment by ajaxery — July 22, 2008

http://www.savethedevelopers.org

Comment by davethieben — July 22, 2008

JustinMeyer:

We should find a way to reward those chosen few who have demonstrated the courage to make the world better for the rest of us.

When you say “the rest of us”, do you mean us developers? Or people generally? Dropping support for the second most common browser platform does not sound like a good way to improve people’s lives.

Comment by deanedwards — July 22, 2008

We’ve already dropped support for it since a year back.
It’s just too horrible..

Comment by Mikael Bergkvist — July 22, 2008

The last place I worked did not allow Firefox because IE6 was “more secure” (translate as “IT can push out security patches” even though said patches often ran 6 months or more behind due to compatibility testing, etc.) Not sure why they “couldn’t” manage Firefox patches just as easily. As developers, we finally lobbied to get Firefox at least installed, but we could only use it inside the firewall for dev/debugging. Any web browsing still had to go through IE. Was rather comical/sad/scary.

Comment by bmoeskau — July 22, 2008

I wonder if it would be possible to import the gecko or webkit engines into the IE frame, effectively keeping IE the same (look-wise), but running a high quality and more standards compliant rendering engine.

Comment by matanlurey — July 22, 2008

>>A friend a mine who used to work for the govt. tried to upgrade over a 1000 PCs to IE7.
.
Gov’ts work (or don’t work) that way. My state gov’t sent me a beautiful 4-color pamphlet (probably cost taxpayers a fortune) on taking care of my child for the first 18 months of its life. All sorts of important health and safety issues. When did they send it? When my child was just over two years old.
.
I called and asked why I got it so late. (It took a LONG time to find someone who knew what I was talking about, but I was fascinated by the system and couldn’t give up my search.) Well, they explained, they were behind in the mailing.
.
I said, here’s what you do. Just skip everyone and start with the people who just had babies today.
.
Oh no, they couldn’t do that. They had to send them to everyone who had babies. That was only fair.
.
So, yeah, gov’ts of levels, local, state, and national, will be running IE6 forever. If the gov’t eventually goes Macintosh in 20 years, they’ll be running IE6 in Parallels.
.
As for myself, through 2008 making sure IE6 works on my site. But every time the user comes to my site, I’ll let them know how much longer I’ll support IE6 and that IE7 will run my (computationally-intensive) site 8 times faster than IE6.
.
My message won’t be accusatory. Merely informational.
.
My local newspaper doesn’t even test on IE6 anymore (and the architect tells me they have 20-25% IE6 users). When a user calls with an IE6 issue, they are told to upgrade to IE7. That’s a lot harsher than my plan. I don’t want to piss off customers–apparently my local newspaper doesn’t give a damn about IE6 users any more.

Comment by Nosredna — July 22, 2008

Web developers and designers should be using progressive enhancement. If designing for IE6 (and earlier) is too taking too much time, then just serve simple HTML to these browsers and support basic functionality. If someone is using IE6, then it’s fair to assume that they don’t need to see the latest js/css enhancements.

Comment by WillPeavy — July 22, 2008

We’ve started doing what Spyke suggests above. We put IE6 in our functional specs with a caveat that the site will function in IE6 (it’s navigable and core features work) but look and feel may not match comps. And some things like animations, hover effects, bells & whistles may not be present. Basically, IE6 will operate like 2004, current browsers will function like 2008. This doesn’t mean we stop supporting IE6 at all. What it means is that we have a means by which we decide to resolve IE6 bugs as wontfix or not.

Comment by tack — July 22, 2008

Considering the general theme of this site is Javascript-related, I don’t think there’s much to celebrate about dropping IE 6. IE 7 made almost no changes to JScript. Until IE’s DOM* is fixed and it gets some basic functionality supported by other browsers, the handful of CSS improvements in IE 7 aren’t that exciting to me.
.
* I’m sure it’s wishful thinking, but designMode/contentEditable improvements would be welcome too!

Comment by eyelidlessness — July 22, 2008

Personally, I’m with Dylan. I think IE6 will be around for at least a year. Here’s something interesting information about how people don’t necessarily upgrade their browsers — http://googleonlinesecurity.blogspot.com/2008/07/are-you-using-latest-web-browser.html.

Comment by skypoet — July 22, 2008

Yeah, how would any of you feel if suddenly the DOT stopped supporting your 1986 Honda Accord, making it illegal for you to drive on the freeway?

You kids in your shiny new 2008 Toyota Camry’s think you know everything don’t you?

Just because you’re too (lazy,inexperienced,incompetent) to deal with older drivers on the public roads isn’t OUR problem — it’s YOURS. Grow up and get over it, you little twerp. Someday your shiny new toyota will be a rusty old rattletrap too. God forbid you ever fall in love with the thing.

;)

Comment by wwwmarty — July 22, 2008

@wwwmarty: You can’t possibly think that spending 1/3 to 1/2 of our time making IE behave correctly (according to standards, as it claims to support “standards mode”) is ideal. Or that we can all afford to do so. Or that your analogy even fits? I mean, IE 6 is the 1986 Honda Accord, sure, but with faulty fuel line connections that cause the vehicle to blow up under normal usage.
.
Nevermind that no one’s “banning” anything, and unlike the car analogy, browser upgrades are free.

Comment by eyelidlessness — July 22, 2008

@wwwmarty/eyelidlessness

Using IE6 is a more a type of ignorance than a social or economic burden. It is equivalent on relying on divine intervention to heal your son’s illness when a spoonful of (modern) antibiotics is the cure.

I argue that if you feel like IE6 “must” be supported, you “must” also support text-based browsers on your fancy ajax sites.

That being said, IE6 support should (like mentioned above) start to be degraded wherever possible. Google and several others are already taking steps toward this, where IE6 users do not receive the same options/effects/interface of the more modern browsers, but still have nearly identical functionality. This is a good standard to have period, as the developer will never know for certain the JavaScript or other restrictions on the client side (intranet type setups withstanding).

Comment by matanlurey — July 22, 2008

It’s easy to say “just give IE6 users a basic view.” But take my local newspaper. They have a crazy CMS, not a static page. It was simply taking up too much time to support IE6 (in their estimation).

Comment by Nosredna — July 22, 2008

I would love to get rid of IE6 in my company, but when some people asked me what the top 10 most relevant bugs in IE6 are, sadly, I was at a loss of words. I know about multiple CSS selectors, PNG alpha, the box model, but that’s basically it.

What ARE the top 10 issues with IE6 for web developers? I’d love to see that list… in descending order of nuisance. Give me firepower, please!

Comment by igitur — July 22, 2008

CSS: support for selectors is very poor, positioning is pretty poor, there are obvious actual bugs rather than omissions (all with fancy names like the 3px jog). I’m pretty sure IE6’s box model, per se, is mostly to spec except with table elements.
.
DOM and Range: oh, the piles and piles of incompatibilities. Basically, look at the compatibility tables on quirksmode.org, they’re the most meticulously collected details I know of.

Comment by eyelidlessness — July 22, 2008

I had a case where I had to rewrite large sections of code (in particular–rearrange loops to fit setTimeout caqlls) because they would time out on IE6. This made the code slower for all browsers.

I also had a visualization where I had to turn off text selection as you dragged the mouse around the screen. All 4 ways I could find that were documented to keep the text from being selected actually slowed IE6 down dramatically so I had to add timing code.

It’s not the bugs that get me. It’s the fact that I can just be cruising along on FF, Opera, and Safari and I then I have to rewrite my code (often lessoning the experience for EVERYONE). It just feels like a terrible waste.

Accomodating IE6 limits the OpenWeb and drives developers to AIR and Flash and Silverlight. IE6 makes all the web programmers in the world less productive.

I’m speaking in terms of RIAs that are competitive with Flash. I agree with supporting IE6 fully on sites that are mostly server-based.

Comment by Nosredna — July 22, 2008

The problem of IE6 lies mainly in the corporate market. A lot of companies developed their intranet applications purely on IE6 (screw standards, we just need it to run in IE6). Now they are stuck with a lot of applications, that just do not run in anything else but IE6, blocking automatic upgrades to IE7 as that would mean a massive investments into their internal application upgrades – basically they would have to contract all the original vendors to update these applications.

Before this problem is settled – and it does not look like it will be settled in a nearby future, we will be stuck supporting IE6.

The saddest part is, that some of these companies still insist on developing for IE6, allowing vendors not to support other browsers such as FF, Opera or IE7 and just deepening the sh*thole they (and we) are stuck in.

Comment by OndraM — July 23, 2008

eyelidless:

“IE 6 is the 1986 Honda Accord, sure, but with faulty fuel line connections that cause the vehicle to blow up under normal usage.”

Your analogy is even worse than mine. When IE6 came out, it was the greatest thing sinced sliced bread. And it’s still just as good as it was then.

No, what you’re saying is that faster, newer cars have been created that burn cleaner higher octane fuels. Which is fine. It’s also why gas stations still offer 4 grades of fuel, and not just one.

Lazy, inexperienced, incompetent. Which one is it?

Comment by wwwmarty — July 23, 2008

>>Lazy, inexperienced, incompetent. Which one is it?
.
IE6 has serious bugs. Some corporations don’t even allow patches to be run on it, meaning large security problems and the garbage collection nightmare aren’t even addressed at all.
.
It’s crap. It’s not a current version. IE8 will be out and people will still hang on to IE6.
.
The only way to get rid of it is for sites to stop supporting it. Hopefully, IT departments will understand that IE6 isn’t working with modern sites, and will say, “Well, keep IE6 for our in-house applications, but install FF3 for anything you have to do on the web.”

Comment by Nosredna — July 23, 2008

wwwmarty:

It’s incopetence on the side of Project Analysts and Project Managers for requesting and designing applications running purely on IE6 with disregard for standards and other browsers as “inferior” at the time. It shows their incompetence to correctly estimate the life-cycle of platform (IE6) with respect to the life-cycle of the required application.

It’s laziness of the developers for developing purely IE6 application, instead of producing standards based applications with IE6 tweaks and exceptions.

Comment by OndraM — July 28, 2008

Leave a comment

You must be logged in to post a comment.