Dealing with IE6

[Nik Coughlin]

Further to my critique request, and out of a wish to mollify as many
naysayers as possible, rather than the "arrogant" technique of informing
users that a site looks bad because they're using an old browser, what would
you say would be the best course of action?

My approach recently has been full support for IE 6 if the client is willing
to pay for it (hey, I invested a lot of time in learning about IE6 bugs,
hacks and workarounds!), otherwise I make sure that the site's not too
broken (you can click links and read text) and give them a message
explaining why it looks crap. I'm considering reconsidering this though

Here's a suggestion to simply not serve IE 6 and less users any stylesheets
at all:
http://simplebits.com/notebook/2009/02/13/iegone.html

Here's a posting that talks about different potential levels of support for
IE, along the one that the author thinks is best (make sure it's not broken
and leave it at that, which is what I've been doing, well, that and the
"arrogant" message):
http://www.friskdesign.com/2009/04/20/why-i-support-ie6/

Do you give the client the choice of supporting IE6 at extra cost? Or do
you just always build it into the price?

Discussion on that here:
http://gadgetopia.com/post/6567

Other ideas or comments?

 

[dorayme]

Further to my critique request, and out of a wish to mollify as many
naysayers as possible, rather than the "arrogant" technique of informing
users that a site looks bad because they're using an old browser, what would
you say would be the best course of action?

What you did was not arrogant. That does not mean it was the best thing
to do. Deliver but try not to explain meta issues.

I would not trouble clients with issues about which it would be merciful
not to trouble them with. And this is one of the issues. Remember that
IE6 is still a sizeable chunk of users, I would not discuss providing
for such a big chunk with a client.

It is not an all or nothing affair. You can just make sure IE6 users see
the material and it is *pleasant enough*. There are no rules! It is a
matter of judgement and I doubt there would be anyone more capable than
you to do this.

If there is a complaint about dissimilar looks, that is the time to
discuss the issue with the client or to tell them that to get the look
*much closer* the same as later browsers, will mean more time and cost
etc. The trick is to make sure that what is still a sizeable chunk of
users gets to see more than a page with "CSS off completely".

 

[rf]

Further to my critique request, and out of a wish to mollify as many
naysayers as possible, rather than the "arrogant" technique of
informing users that a site looks bad because they're using an old
browser, what would you say would be the best course of action?

My approach recently has been full support for IE 6 if the client is
willing to pay for it (hey, I invested a lot of time in learning
about IE6 bugs, hacks and workarounds!), otherwise I make sure that
the site's not too broken (you can click links and read text) and
give them a message explaining why it looks crap. I'm considering
reconsidering this though

I would say that your site looks just as good in IE6 as it does in other
browsers, much better than, say, IE5.5. The round corners are missing of
course but so are they in IE7.

Here's a suggestion to simply not serve IE 6 and less users any
stylesheets at all:
http://simplebits.com/notebook/2009/02/13/iegone.html

Not a very good suggestion IMHO. Why *intentionally* cripple the site for
14% (w3schools) of your viewers? Graceful degradation is what is needed, not
total annihilation

No matter what you do or say (including arrogant messages ) you will
*not* cause people to install another browser just to view your site. I know
people out there who don't even know one can install browsers. And there are
plenty of other sites out there to view. Serve up something without CSS and
you just drive those 14% of viewers elsewhere. BTW have you actually looked
at your site without CSS? Usable but not all that pretty

Here's a posting that talks about different potential levels of
support for IE, along the one that the author thinks is best (make
sure it's not broken and leave it at that, which is what I've been
doing, well, that and the "arrogant" message):
http://www.friskdesign.com/2009/04/20/why-i-support-ie6/

My approach is point 4 on that page. If it works and looks good enough,
possibly with a couple of IE6 specific CSS rules, then fine.

I most certainly stay right away from point 5, "make your site look exactly
the same". It's not going to look the same in all browsers anyway.

Do you give the client the choice of supporting IE6 at extra cost? Or do
you just always build it into the price?

I don't give my clients any choice. My sites work in all current browsers,
and like it or not IE6 is still a current browser. When usage drops to less
than .5% (as IE5.5 did a year ago) I might consider it to be not current.
Then again I still actually check that nothing totally disasterous happens
even with IE5.5. Your site works with IE5.5. Might not look quite as nice,
but it works. And the three IE5.5 users out there are used to things looking
a little odd.

My sites might not have full eye-candy functionally in IE6 (hover effects on
things other than a elements with a href, round corners) but they work and
are usable. Yes, even on lynx and mobile phones.

I find that the time I need to spend "making it work" in IE6 is very small,
maybe one percent of the total development time. Possibly my style of coding
leads me to not use things that really annoy IE6 (or IE5.5 for that matter).

Of course most of the development time is spent on the server side smarts
anyway. The presentional part (displaying it on a browser) is less than
half.

 

[Andy Dingley]

Further to my critique request, and out of a wish to mollify as many
naysayers as possible, rather than the "arrogant" technique of informing
users that a site looks bad because they're using an old browser, what would
you say would be the best course of action?

Your "IE6 sucks" message is fine. IE6 _does_ suck. There are also
rational upgrade paths away from it, for pretty much all users.

Caveats:

* Don't make the message excessive.

* Make the messsage disappear over time. Client-side JavaScript / week-
long cookie will show it once per week for that user and that's
plenty. This is also a good example of something where graceful
fallback for non-JS users is entirely aceptable. If they're still
using IE6, they might have a good reason that they have to. Don't rub
their noses in it.

* Maintain the message. Don't put up any messages that will begin to
look ridiculous in less than the site's lifetime. Saying "Your IE6 is
old" is OK, suggesting an "Upgrade to IE7" already looks dated and so
will "Upgrade to IE8" or IE9 within a few years, when IE11 is around
and renamed "Encarta 11"

* Make the message accurate. Don't serve "Your IE is old, upgrade to a
newer IE" message _unless_ you're certain that they're not a Unix
user. Or else they're running BSD, in which case they see it as a form
of flattery and their superiority over the feeble Windows lusers.
Certainly don't trust browser-sniff strings.

 

[Beauregard T. Shagnasty]

I most certainly stay right away from point 5, "make your site look
exactly the same". It's not going to look the same in all browsers
anyway.

That is a very good point.

One thing a lot of people asking for critique seem to oft forget, is
that the masses will never look at a site with more than the one browser
they use. Only us developers do that.

And, if the number of commonfolk who ask the question in the anti-virus
groups, "what is the best a-v for Windows 98?" is any guideline, IE6 is
going to be around for a long time.

 

[dorayme]

If they're still
using IE6, they might have a good reason that they have to.

Some a government departments still use IE6 because their IT people fear
a flood of support problems if this changes; inertia helped along by
fear of work.

 

[Jonathan N. Little]

Some a government departments still use IE6 because their IT people fear
a flood of support problems if this changes; inertia helped along by
fear of work.

As I have said before, take a quick glance at the monitor of your bank
teller. You might be surprised how many are still running Windows 2000...

 

[Andy Dingley]

As I have said before, take a quick glance at the monitor of your bank
teller. You might be surprised how many are still running Windows 2000...

Or OS/2

Or WIndows/NT

Saw one just this weekend!

 

[Jonathan N. Little]

Or OS/2

Or WIndows/NT

Saw one just this weekend!

Scary! Remember what a PITA it was to change a driver in NT! Not there
is an OS that you gotta love.