Why we don't support Internet Explorer (2009)(engineering.silkapp.com)
engineering.silkapp.com
Why we don't support Internet Explorer (2009)
http://engineering.silkapp.com/post/32258824729/how-i-learned-to-stop-worrying-and-love-web-development
11 comments
I have an odd feeling that I've commented about this one before. Maybe this article cycles back around to HN every year or two?
There are large populations of users who are forced to use IE6. Still. In 2012. They are not allowed to upgrade their work computers even though they may hate IE6 and use a modern browser at home.
The reason is usually that their corporation, hospital, organization, agency, etc. paid lots and lots of money many years ago for custom software that cannot run on newer versions of IE, or is not certified to be reliable/secure/accurate/etc. on newer browsers, and so those organizations must either force their users to use IE6, or pay lots and lots of money, again, to replace/re-certify custom software that is still working perfectly well.
So the business case for forcing your employees to stick with IE6 is still very rational in some places, even today.
If you're developing a new way to tweet facebook likes or something of that ilk, that's just fine; anyone using IE6 in your userbase should be encouraged to stop.
But if your users might include any of the sorry souls who are forced to use IE6, at the very least you should be kind when you tell them they aren't going to be able to use your site.
There are large populations of users who are forced to use IE6. Still. In 2012. They are not allowed to upgrade their work computers even though they may hate IE6 and use a modern browser at home.
The reason is usually that their corporation, hospital, organization, agency, etc. paid lots and lots of money many years ago for custom software that cannot run on newer versions of IE, or is not certified to be reliable/secure/accurate/etc. on newer browsers, and so those organizations must either force their users to use IE6, or pay lots and lots of money, again, to replace/re-certify custom software that is still working perfectly well.
So the business case for forcing your employees to stick with IE6 is still very rational in some places, even today.
If you're developing a new way to tweet facebook likes or something of that ilk, that's just fine; anyone using IE6 in your userbase should be encouraged to stop.
But if your users might include any of the sorry souls who are forced to use IE6, at the very least you should be kind when you tell them they aren't going to be able to use your site.
tldr: Ignore this article. It's 3+ years old and the marketplace has changed significantly. IE6 & 7 should no longer be on anyone's support matrix. IE8 is heading out, IE9 was really damn good, and IE10 (being released this month) is as fast or better than Chrome and has very good support for most html5 and js features. Basically, with IE10 Internet Explorer is going to be on par and perhaps even start pushing ahead of other browsers.
Lemme also say that I'm not a Windows user, Chrome is my fav, etc. But from the testing I've done and people I've talked to, that is what I base my assessment on.
Lemme also say that I'm not a Windows user, Chrome is my fav, etc. But from the testing I've done and people I've talked to, that is what I base my assessment on.
I used to rail against IE. I even bought 2 domains that I planned to use as a totally cliche "IE sucks" website. In the past six months I've changed my attitude towards IE. It still leaves much to he desired and versions 6 - 8 are still worth complaining about but I must say the IE9 has been a huge improvement to the point where I've stopped my grumbling. The amount of work it takes to get JavaScript and CSS working consistently across browsers is still more with IE9 but the amount of extra effort is so trivial that I think what I perceive to be harder to get to work in IE compared to Gecko and Webkit is actually an illusion based on a bias towards hating/loving to hate Internet Explorer pre-IE9 and that it's actually not all that difficult at all. At this point I think my (and many other) developers' disdain for IE says more about my own skill at the time than it said about Internet Explorer. And even now when I have a troubling time with IE and curse it to hell I think it's only difficult for me because I've invested so much time into learning the quirks of Gecko and Webkit but always ignored/avoided learning about Trident and IE because of my prejudice. Many times I had a case of "my code is right and this browser is wrong no matter what".
Now I have to say I was wrong about IE. It's not the awful monstrosity of a browser I've learned (or rather been taught as the case many times was) to hate. There's no disputing the fact that Internet Explorer from v6 to v8 got stale, held the web back, and was a pain to work with. The IE team got so used to creating the rules rather than playing by them that they seemed to think their reign would last forever and they seemed to ignore all standards for a time (in the case of the Trident CSS box model I have to admit I think that makes more sense). However this is not then. Today's Internet Explorer has a bigger problem with prejudice and it's reputation than anything technically inferior.
There's still a lot left to be desired from IE like a shorter release cycle that adopts new specs and standards like the Webkit and Gecko browsers are doing but despite that there's still not very much you can do in Firefox, Chrome, Safari, or Opera that can't be done with roughly equal effort in IE. And with IE10 the gap is closing even more. Sure, Microsoft has made some intentional decisions to support or not support certain features or purposely implement them differently but both Firefox and Chrome make similar decisions and are applauded for them or at least don't get chewed out by the techie crowd for it. I've learned that IE users as a whole are actually a different kind of user altogether with different needs and competency levels than are users of other browsers and I'm talking about those users as a whole, not just the ones working on corporate networks. That considered, I think IE has done a great job of walking the line between keeping up with the pace of the web and serving those users.
This is all coming from a reformed IE hater. I learned that in most cases the browser wasn't the problem, it was my laziness or lack of skill that was. That was hard to admit. To all the people thinking of dropping support for IE or hating on it as its so easy and trendy to do, I'd tell you to really give it a shot and learn how to support it. You'll be a better developer for it and you'll find it's not as bad as it was when you swore it off all those years ago.
Edit: Spelling and brevity (tried my best to shorten this, sorry)
Now I have to say I was wrong about IE. It's not the awful monstrosity of a browser I've learned (or rather been taught as the case many times was) to hate. There's no disputing the fact that Internet Explorer from v6 to v8 got stale, held the web back, and was a pain to work with. The IE team got so used to creating the rules rather than playing by them that they seemed to think their reign would last forever and they seemed to ignore all standards for a time (in the case of the Trident CSS box model I have to admit I think that makes more sense). However this is not then. Today's Internet Explorer has a bigger problem with prejudice and it's reputation than anything technically inferior.
There's still a lot left to be desired from IE like a shorter release cycle that adopts new specs and standards like the Webkit and Gecko browsers are doing but despite that there's still not very much you can do in Firefox, Chrome, Safari, or Opera that can't be done with roughly equal effort in IE. And with IE10 the gap is closing even more. Sure, Microsoft has made some intentional decisions to support or not support certain features or purposely implement them differently but both Firefox and Chrome make similar decisions and are applauded for them or at least don't get chewed out by the techie crowd for it. I've learned that IE users as a whole are actually a different kind of user altogether with different needs and competency levels than are users of other browsers and I'm talking about those users as a whole, not just the ones working on corporate networks. That considered, I think IE has done a great job of walking the line between keeping up with the pace of the web and serving those users.
This is all coming from a reformed IE hater. I learned that in most cases the browser wasn't the problem, it was my laziness or lack of skill that was. That was hard to admit. To all the people thinking of dropping support for IE or hating on it as its so easy and trendy to do, I'd tell you to really give it a shot and learn how to support it. You'll be a better developer for it and you'll find it's not as bad as it was when you swore it off all those years ago.
Edit: Spelling and brevity (tried my best to shorten this, sorry)
I've been rumaging over this same thought myself and I think there's a real key point here that is sort of distinguished between in the article, mainly it boils down to just knowing your audience, but on a broader level I think its this: are you developing a web app or a website? For me, if I'm building an app I've also gotten to the point where I just ignore IE completely, mainly for reasons stated in the article (writing tons of work arounds that convolute the codebase, absolute lack of support for MANY features, etc). Also I think in most cases( it does depend) for most apps and their audience, very few are going to be using IE, so I don't really care about losing that 1% of my target group, especially if it helps me build a better app. For websites, sadly, the story is in most cases different and the majority of users will come from some version of IE. (Although, personally I've notice this changing a lot in the last year for my sites, we've been getting huge spikes in mobile - which yes, means no IE, so I think unless microsoft does something to take over mobile in 5 years devs aren't gonna have to worry about it.
"devs aren't gonna have to worry about it"
Yeah, out of context but I think this is telling. I know a lot of devs who ignore one browser or another or ignore mile or ignore desktop or only do responsive design as opposed to building out a mobile only site and vice versa. In most cases they do some mental gymnastics to convince themselves they don't need to worry about a certain segment of the market. I think is is dangerous in that it can cost you a lot of money. Many times these decisions are made because they don't want to pay for supporting X or don't want to learn how. Then they justify it by saying that market segment X is too small to support. Instead of being based on reality it instead is just a self-fulfilling prophecy. Market segment X doesn't use site Y not because they're so small but simply because X is not supported. Imagine HN launched today. Today is HN's first day and for some reason PG said we won't support IE because HNers are all using "better" "modern" browsers. Then he checks the logs and sees a very small portion of IE users that steadily shrinks. Is the portion small because Hackers who use IE are few and far between? Maybe. Or maybe it's because people who visited HN on IE couldn't access the site and so they didn't spread word to their friends who may also use IE and of course they never return because the site never worked for them to begin with and so the number of IE users remains small enough to be ignored. But in reality we know that there are definitely enough HN users on IE to justify supporting it. I know because I see my own logs from HN referrals from posts I've made in the past or people reading my profile. The percentage is lower than other browsers but definitely more than a trickle and surely not low enough to justify not supporting IE if I ever had to make that decision.
My whole point is that I think it's dangerous to make assumptions about your audience because they often become self-fufilling prophecies. You can't say a certain platform isn't worth supporting until you have data to show it and you can't get that data (at least not accurate data) unless you support the platform in question to begin with.
I do agree with you about Windows phone though. At this point, at least in the mobile space, I think companies would be smart to make their mobile browsers compatible with what's out there now rather than releasing a mobile browser and expecting developers to cater to it. I wonder if the iPhone would've taken off like it did had they decided to release a very stripped down version of Safari and expect devs to modify their sites to support it.
Yeah, out of context but I think this is telling. I know a lot of devs who ignore one browser or another or ignore mile or ignore desktop or only do responsive design as opposed to building out a mobile only site and vice versa. In most cases they do some mental gymnastics to convince themselves they don't need to worry about a certain segment of the market. I think is is dangerous in that it can cost you a lot of money. Many times these decisions are made because they don't want to pay for supporting X or don't want to learn how. Then they justify it by saying that market segment X is too small to support. Instead of being based on reality it instead is just a self-fulfilling prophecy. Market segment X doesn't use site Y not because they're so small but simply because X is not supported. Imagine HN launched today. Today is HN's first day and for some reason PG said we won't support IE because HNers are all using "better" "modern" browsers. Then he checks the logs and sees a very small portion of IE users that steadily shrinks. Is the portion small because Hackers who use IE are few and far between? Maybe. Or maybe it's because people who visited HN on IE couldn't access the site and so they didn't spread word to their friends who may also use IE and of course they never return because the site never worked for them to begin with and so the number of IE users remains small enough to be ignored. But in reality we know that there are definitely enough HN users on IE to justify supporting it. I know because I see my own logs from HN referrals from posts I've made in the past or people reading my profile. The percentage is lower than other browsers but definitely more than a trickle and surely not low enough to justify not supporting IE if I ever had to make that decision.
My whole point is that I think it's dangerous to make assumptions about your audience because they often become self-fufilling prophecies. You can't say a certain platform isn't worth supporting until you have data to show it and you can't get that data (at least not accurate data) unless you support the platform in question to begin with.
I do agree with you about Windows phone though. At this point, at least in the mobile space, I think companies would be smart to make their mobile browsers compatible with what's out there now rather than releasing a mobile browser and expecting developers to cater to it. I wonder if the iPhone would've taken off like it did had they decided to release a very stripped down version of Safari and expect devs to modify their sites to support it.
You guys are so cool.
Whilst it's easy to assume that your tech-savy audience won't be using IE for their desktops, please remember that anyone with a Windows phone will also be forced to use IE. I've used a Windows phone, and it's not as dreadful as you'd think. In fact it's quite a nice experience. As tempting as it is, please don't start completely dropping support for IE unless that extra dev time really wouldn't affect your customers (considering your future customers too).
Even if Windows phone is a viable platform only a very small minority of users actually use it. As a startup there are more important things to do than optimizing your product for a platform that no-one uses. If you as a user want to use those sites just use a phone/browser that supports current Web standards.
ie9 mobile supports most modern standards, are you planning on not supporting firefox mobile/opera mobile either?
This dates back to Oct 2009, just after IE8 release.
I can't figure out if they're saying they're dropping support for IE6, or ALL of IE.
If the latter, give me a break. By my count, over 50% of our traffic comes from IE.
Extra development to support IE7 and above last year: 2 hours.
You're telling me it's not worth 2 hours per year to stop yourself from alienating 50% of your market?
Plus, IE8 and IE9 are pretty compatible in that whatever works in Webkit/Firefox will pretty much work every time in IE8+.
If the latter, give me a break. By my count, over 50% of our traffic comes from IE.
Extra development to support IE7 and above last year: 2 hours.
You're telling me it's not worth 2 hours per year to stop yourself from alienating 50% of your market?
Plus, IE8 and IE9 are pretty compatible in that whatever works in Webkit/Firefox will pretty much work every time in IE8+.
> Plus, IE8 and IE9 are pretty compatible in that whatever works in Webkit/Firefox will pretty much work every time in IE8+
So this tells me that explicitly testing in IE8+ is a waste of time, then.
So this tells me that explicitly testing in IE8+ is a waste of time, then.
The key phrase in my comment was "pretty much." You should still test, but there's been virtually no time spent making something work in IE8 that works elsewhere.
Sure, if you're comfortable with a website that will "pretty much work" for a decently-sized chunk (YMMV) of your users.
If you'd rather have it "definitely work" for those users, then test it.
If you'd rather have it "definitely work" for those users, then test it.
(2009)
The points this article makes no longer apply.
Unless you need webgl, appcache or contenteditable there's no good reason not to support at least IE9. The effort to support it is negligible, if you're already doing opera, firefox and webkit. And if you're doing contenteditable on all three of those, you might as well do IE as well and make those gray hairs actually mean something.
Unless you need webgl, appcache or contenteditable there's no good reason not to support at least IE9. The effort to support it is negligible, if you're already doing opera, firefox and webkit. And if you're doing contenteditable on all three of those, you might as well do IE as well and make those gray hairs actually mean something.
> Google Wave team also chose to drop Internet Explorer support
3 years is a significantly long period of time in the browser world and Chrome was a 1 yr old browser and there was no IE9. Also, there was no windows phone 7.
Once in a few months, an article pops up telling how a startup saved $100k by not supporting IE or how hundreds of developer hours were saved by not caring about the things that IE failed to render properly. It narrows down to the number of users an app or service has and also the demographics of the users. If you are building a hipster social network or a video post processing tool, ignoring IE is going to save you significant time since it can be assumed that a good percentage of the users are going to be on macs using the latest browsers. On the other hand, if you are catering to e-commerce, finance or any other common industry even thinking about ignoring IE isn't wise at this point. With jQuery and graceful degradation, fixing the major issues shouldn't be really hard.
PS: Can just-launched-starup ignore IE for their app? Probably.
Can Pinterest ignore IE8? Of course not. Even if it is 10% of their user base (I bet it will way more), they will be ignoring a few hundred thousand or may be a few million users!