Iris Classon
Iris Classon - In Love with Code

‘Stupid’ Question 28: Can devs refuse to create a web app that works in IE for clients?

[To celebrate my first year of programming I will ask a ‘stupid’ questions daily on my blog for a year, to make sure I learn at least 365 new things during my second year as a developer]

Can we refuse to develop for IE?

I was really curious to see what other developers had to say about this. Cross browser compatibility is important, and can be a challenge. Our clients use different browsers to access the applications we make. And some clients don’t even update their browsers, but we update our technology which adds to the problem. IE has been a favorite punching bag as it supposedly offers the biggest challenges, and the most problems. And I’ve learned that a few developers outright refuse to make applications that work in all IE browsers, or set a limit to only support IE version 7 and up. But can we do that? Can we refuse?

Greate article by Lea Verou on the subject

Greate article by Lea Verou on the subject

I again asked the question on Twitter and Facebook and got the following replies:

No. If the business case states that the target users are using IE, it is a fatal blow to refuse a technology used by the target group. This is one of the hard things with being a passionate developer, some times business comes in the way for the “right” tech. After all, its the business that will pay your salary in the end.

I got a new wonderful client that accepts the app we are building will only work in browsers that support WebGL (come on IE you can do it!). I’m so happy we didn’t have to resort to Flash or Java in this case.

We can turn down the project, but we can’t refuse IE during a project. That said we should help the client make the right decision. It all depends on what we’re building and if we have any stats what browsers their users are using.

We can’t all do revoke support for <IE9 like 37 Signals

Its an idealistic developers dream to do so. In the real world where business rules, if the website is supposed to target the general public, or the company itself has IE as a standardbrowser. As a techie you just cant refuse. As I said before, Technology is never the answer, it is the tool.

The thing is, majority of users are still using IE. It’s not a “marginal” browser. Its usually the default browser. So even if WebKit feels like a better dev enviornment, reality is that people will mostly use IE to access the web.

I think a better question is “What versions of IE should we recommend our customers to support?” .

oh I know a few horror stories down that alley :P

if it’s a public web site it’s only reasonable to support all major browsers: IE7+ (maybe 8) and current versions of all others

would the revenue lost for not serving legacy customers justify the extra development & maintainance cost? Sorry, a biz answer!

it’s often best to set aside some time to “educate” the customer on the minefields

devs should just suck it up and deliver what users/business want and not be dramaqueens

do you mean refuse to make something that ONLY works in IE?

sure. I know people who add IE support on as a separate line item, and is therefore negotiable. At this point, IE6 support in particular is just flat out, and some people charge lots extra for it.

Depends on the kind of app and target users, I think. Many large companies and gov departments still use old versions of IE. At least here in Australia that’s true, not sure how it compares to corporate & government browser use other countries.

clients provide specs/reqs. If some of them look unreasonable to you try to convince the clients or you don’t take the contract

Yes, they can. They ought to be instantly out of a job if they do, but they surely can. The customer is always right.

taking the job on their conditions and then not delivering is a no-go

kogan decided to impose taxes on some IE7 usershttp://www.lifehacker.com.au/2012/06/kogan-imposing-tax-on-shoppers-who-use-ie7/ …

the customer is often right, but not always …but they do decide in the end if they want to pay

They can; but it may limit the amount of customers they can reach. I’m not sure why a dev would consider this as an option

A responsible developer will work with the client to determine how important legacy browser support is.

Also, be clear that supporting, say, IE7 requires extra work for which the client will be expected to pa

currently, our frontender is creating stuff for Chrome, Fx, Safari, IE8+ first and then fixes/degrades to IE7 afterwards

It depends. Does it add a significant amount of work? If so, I would. If not, then not.

then you charge them extra. Specifics should always be part of a contract. But that’s not a dev’s job to sort out.

clients provide specs/reqs. If some of them look unreasonable to you try to convince the clients or you don’t take the contract

if they don’t know the implications, then the consultant have failed in doing a good job for the customer tbh

advise them of approx %age using IE. build without testing, charge additional fee to add compatibility later.

tech advisor that sits in on the spec of the assignment

I’d recommend charging extra for IE 7 support. Supporting IE 7 can add quite a bit of time to a project.

Boils down to “can I afford to lose customers over it vs how much is the extra effort worth to me as a business” … hardly a technical decision if you ask me.

need to make sure they know they’re cutting off large numbers of users, then get them to sign it off.

IE 8 on up are fine for most mainstream development scenarios in my experience. Each supported browser adds QA time, however.

Comments

Leave a comment below, or by email.
Michael Malura
8/21/2012 2:41:37 AM
The Brwoserproblem is really annoying. In our Company Project i have to care for the IE. 
Mark
8/21/2012 2:50:31 AM
If anyone refuses let me know, ill take over your job for you and deliver what the client wants, he pays my salary after all 
Fredrik Mörk
8/21/2012 3:12:12 AM
You can refuse or push through whatever you want, but then also take the consequence. If the client demands IE 6 support, and you have the choice of doing that or finding another job, well that's up to you, right? 

However, as professional software engineers it is our obligation to make the clients aware of what they are really asking for, which is not always the case. Are they aware of the security implications on sticking with old browsers and OS'es? Also, I read a lot about pixel-perfect design. This is also a matter of making a reasonable investment. Is it really worth going the extra mile(s) to have a pixel-perfect design in IE6, IE7, IE8, Chrome. And what about Firefox, who has release versions 10-14 during this year?

I sometimes feel as too much attention is given to detail that really will make a very minor difference for the user experience.

So, I can see how web devs would like to refuse to have for instance IE6 in the browser-version-brew, but in the end of the day, the client pays the bills and puts the food on your table. It's a balance between using your knowledge to make the client aware of their demands, to help guiding them to good decisions, and perhaps letting go of some of your principles to reach some common ground. 
Richard P
8/21/2012 4:28:00 AM
I personally follow Google in what browsers they drop support for - if your browser is old enough that Googles suite of apps aren't guaranteed to work in it, then I won't develop for it either.

Google don't support IE versions lower than IE 8 at the moment... 
David Corbett
8/21/2012 5:05:43 AM
I suppose it's the same problem that many other professions/trades have: eg. Architects/Builders of houses usually prefer to work on a green-field new project as against extending an old one with delapidated plumbing, electrical wiring, heating systems etc. The new one is probably a more creative experience, can be done for less cost, with less compromises and less frustration. 

Eventually, there comes a time (if there isn't any inherent value in retaining the old ediface) when a knock-down and rebuild becomes more cost effective in the long term. 

I suppose it's part of the developer's job to know when that time comes and to be able to make the case for it when it does. 
Mark
8/21/2012 5:23:24 AM
Reply to: David Corbett
David , i have found that its actually the opposite, depends on how lucky you are. There are some people that ABSOLUTELY refuse to change. Ive done , very recently work on a CLASSIC asp website , and that goes back how far? 20 yrs? ??? and their main application is still on COBOL on old MAINFRAMES yes the old ones with the TAPES , now even the new mini ones. 

They say it worked , still works and that's what they want, either you make it work for them or they find someone else who will :( 
David Corbett
8/21/2012 6:42:28 AM
Hi Mark,

And that's a testimony to that system and your ability to make sense of it and modify it.

I'm not advocating change for change sake. But, everything around that company is changing. The cost of maintaining that system is rising, the cost of finding people who can maintain it will rise, the cost of being unable to easily incorporate recent technological advances into the system will rise. These rising costs will eventually create enough competitive pressure for a re-think.

So, for me, it looks like the developers need to know a lot. They need to know about the old and the new and how to bridge the gap between them. They may also have a role in advocating when is the right time to make the change? 
Sascha
8/21/2012 7:18:48 AM
Reply to: Mark
I know what you mean, Mark.

I am working for an insurance company in germany where most of the applications are written in COBOL some also with IMS-DB system (for those who don't know it - IMS-DB is a hierarchical database). Some applications have nearly the same age-group like me (1981) and they still do their job :-)
Many COBOL developer in my company doesn't have the heart to do more than bug fixing on these old applications because they don't know what happen :-) 
And those old COBOL applications do their job more reliable and sometimes faster than the new applications. Let me tell you that after the rollout of SAP many nightly batch jobs are significant slower than the old ones based on COBOL.

Sorry I wandered from the subject :-)

Like Mark said: if the customer really wants IE 6 support and he can constitute why it should be implemented, then you have to implement it, if you want the job. 
But if the reason for implementing IE6 is not conclusive then you can try to persuade the customer with the best reason ever: The development becomes cheaper if IE6 support is not implemented (in my company this argument is the ace in the hole ;-) ) 
Jon
8/21/2012 8:25:35 AM
"We decided that in order to get a viable product out the door we would simply not support IE. It was possibly the best decision 4ormat ever made." http://techcrunch.com/2012/04/01/bootstrapped-startup-saves-over-100k-by-dropping-ie/ 
Mark
8/21/2012 11:00:38 AM
Reply to: Jon
Jon, this from your link , "almost three years after launching 4ormat, not a single person has ever contacted us requesting support for Internet Explorer".

thats becuase , they like me would have just gone away, not worth the bother to complain

You said you aimed at creative people?, they/we/I/Creative types don not take kindly to being TOLD what we can and cannot use :) 
Rohit
8/27/2012 11:11:09 AM
In India as well, IE is like de-facto standard of Government organizations. Website must work well on IE no matter it fails on other browsers. 
Ladislau Radu Nagy
1/21/2013 11:10:01 AM
I remember a great lesson learned from FoxIE a small video series from channel 9.
http://channel9.msdn.com/Series/Foxie
Develop web applications based on the features supported by the browsers, don't target specific browsers or devices.
I think this is the reason why javascript libraries like modernizr, jquery and others exist.
If the browser is not supporting the cool interface, funky controls and other good stuff that your web application can do, then offer only the basic html features, don't force it. Gmail and Yahoo mail still have the basic html app for this reason.
Costumers usually upgrade if they see a beneficial reason, don't forget that not everyone is a big company, they are always looking to cut expenses.
Pardon me if I'm off topic with this one, but I don't see what's all the fuss about IE. It's a native OS application, usually outperforms 3rd party apps, performance issues rise when it's cluttered with add-ins and toolbars. Older versions lack todays features, but the same is true about the other browsers. 
BBobob
11/5/2013 6:22:21 AM
You should always be honest about price with customers, so: WebKit & Mozilla £100; add IE10 £200; add IE9 £300; add IE6 £1000. After all it is development time that dictates the cost. 


Last modified on 2012-08-20

comments powered by Disqus