views:

430

answers:

11

The question is simple really, when is it time to finally pull the plug on a browser version and just not care about supporting it any more? As can probably be expected this is in regards to IE6 specifically right now. A site I'm working on right now has a mostly-finished design that needs to be integrated but most of the back-end functionality isn't expected to be done for another 6 or 7 months. Is it really worth fighting with IE6 as apposed to just skipping it and supporting IE7+ instead? Also, the site in question is targeted specifically at photographers, many of which I know for a fact are mac-heads.

+13  A: 

There's no magic answer to this question of course, though like the question the answer is simple. :)

You need to look at the data for visitors on that specific site, and then make a cost vs. benefit analysis, based on the extra effort it will take to support that browser vs. the number of visitors you expect to be using it.

There are other things you can do to make your life easier however. For example:

  • Use frameworks (eg jQuery) to abstract away browser differences and bugs.
  • Don't worry about pixel perfection cross-browser.
  • Send IE 6 a simpler style-sheet, or no style-sheet at all.

Your still supporting the browser in these cases, just not giving it the level of support that you might a modern standards compliant browser.

See Yahoo's Graded Browser Support policy.

Andy Hume
A great answer - your site doesn't have to look *the same* in all browsers - it just has to work!
Ross
+1  A: 

It depends on your user base, check what they're using.

However if it's the general public, I would imagine ie6 will be around for years.

Bravax
+3  A: 

The problem with IE6 it's that it's still the standard browser on Windows XP in quite a lot of corporate environments. So basically Microsoft has to release a credible replacement for Windows XP before we can be rid of the dreaded IE6.

Edit: I had a once had a bug report "on behalf of 27,000 users" (IE6), we had a bug that was making our site unusable for a large corporate customer. In an otherwise slow-moving environment this bug was fixed and deployed in 2 days.

krosenvold
+1 Working for a company where they *just now* start to deploy Windows XP (with IE7, though). We'll not be IE6-free for another year, that's my guess.
Tomalak
Tying IE7 to Windows Genuine Disadvantage hasn't helped either.
Peter
A: 

You can for the large part start ignoring it now if you use a framework such as jquery which abstracts away the differences between modern browser DOM implementations.

DSO
You still have CSS issues to deal with, as well as bugs jQuery can't dodge by itself (mainly involving CSS effects).
strager
+3  A: 

You do not need to lose your head on style sheet bugs and peculiarities of every web browser - no browser's display/layout engine is perfect. It is alright if your web site does not look optimal on some lesser used web browsers which are difficult to support. However, your web pages should be accessible and usable on old web browsers too.

This means:

  • Write valid, well structured HTML, in proper markup that makes sense and is legible without stylesheets too (yes, not all browsers support them, and you never know how your page will be presented anyway). Your pages should make sense with stylesheets disabled.

  • Try to write stylesheets that are as simple and structured as possible given the desired effect. That will increase the likely consistency of the layout and presentation among different implementations.

  • Test your pages with stylesheets on web browsers and setups that you expect to be commonly used by the majority of your target audience, to see that they look good. Test them on a few older and less common browsers, to make sure that they look acceptable and are still legible and accessible, even if the fonts, spacing, and layout is not exactly what you want.

If it doesn't look really good as you designed on a very common browser (e.g. Internet Explorer 7 and Firefox 3 nowadays, and more depending on your audience), that's a shame - even if it's because of a browser bug. Find ways to fix it or work around the bugs, occasionally you'll need to use some tricks to work around such bugs.

I do not know exactly how common various versions of Internet Explorer are. However, in general, if your site just plainly does not work or looks all garbled up on Internet Explorer 6 or a lesser common browser, that's bad.

Tom Alsberg
+8  A: 
Scott Evernden
Do you have a link to that chart?
Andy Hume
Lies! Opera takes less cost than Firefox! (Opera fanboy here. =]) Great data. Would like a link to the original research, though.
strager
http://ejohn.org/blog/the-browsers-of-2009/
Jasper Bekkers
A: 

You've probably got another year or so to support it - Microsoft says service pack support will end on 13-Jul-2010, so if you support it until then, you'll be fine - after that, I think you can happily tell people its nothing to do with you, blame MS :-)

gbjbaanb
When Paypal announced they would block some browsers last year they revealed some people were using ***IE3*** so don't expect it to completely go away. :) But yes, 1-2 more years before IE7 becomes the minimum I think.
cletus
A: 

It sounds like your site does not target the millions of captive IE6 users who work for large corporations, so they may not be a major concern to you.

But for those hoping to reach those users, understand that many large organizations will not be migrating from IE6 any time soon. It is just too costly. The typical user there is not an administrator on their machine and can't change browsers (say, by downloading Firefox or Chrome).

Here's one of the problems with the IE6 to IE7 migration: IE7 is picky about some tags. For example, you can't have a self-closing div or script tag. If you try that, IE7 renders a blank page. In addition, IE7 rejects duplicate attributes in a tag. This is a change from IE6. So, there's lots of sloppily-written intranet apps out there that would have to be fixed before they will work in IE6. And it's very hard to search through the code to find these issues. There's little ROI in doing it, so it just isn't being done.

DOK
A: 

The only thing I would say is to make it clear to your users that you're not supporting IE 6 if you decide not to support it. Your users on IE 6 will probably be annoyed with you, but at least they will know how to make it work. And better still, they're less likely to think that you're just incompetent.

Jason Baker
A: 

All depends on who your target users are and if you can push/dictate a set of browsers on them. If it is an internal or SaaS based model this can be an easier sell, if its a public site then you need to hit a wider base, but it still comes down to the target audience. If your targetting Windows developers for example, supporting old versions of Safari may be a non issue...

schooner
+1  A: 

As others have noted, this question is very subjective, and site audiences differ. That having been said, I use the following as a general guideline:

  1. Support the latest versions of all major browsers (Internet Explorer, Firefox, Safari, Opera, Chrome).

  2. Try support the browsers used by 98% of the browsing public as told by various statistics websites.

Joshua Carmody