Keep in mind that many old-IE users are there because their corporate/gov
entity requires it. Our entire univeristy health/hospital complex, for
example, was on IE6 until...last year, maybe?... because they had several
critical pieces of software written as active-x components that only ran in
IE6. Which, sure, you can say that's dumb (because it is), but at the same
time we couldn't have a setup that made it hard for the doctors
and researchers use the library.


On Tue, Feb 19, 2013 at 10:22 AM, Michael Schofield <[log in to unmask]>wrote:

> Hi everyone,
>
> I'm having a change of heart.
>
> It is kind of sacrilegious, especially if you-like me-evangelize
> mobile-first, progressively enhanced web design, to  throw alerts when
> users hit your site using IE7 / IE8 that encourage upgrading or changing
> browsers. Especially in libraries which are legally and morally mandated to
> be the pinnacle of accessibility, your website should - er, ideally - be
> functional in every browser. That's certainly what I say when I give a talk.
>
> But you know what? I'm kind of starting to not care. I understand that
> patrons blah blah might not blah blah have access to anything but IE7 or
> IE8 - but, you know, if they're on anything other than Windows 95 that
> isn't true.
>
>
> *         Using Old IE makes you REALLY vulnerable to malicious software.
>
> *         Spriting IEs that don't support gradients, background size, CSS
> shapes, etc. and spinning-up IE friendly stylesheets (which, admittedly, is
> REALLY easy to do with Modernizr and SASS) can be a time-sink, which I am
> starting to think is more of a disservice to the tax- and tuition-payers
> that pad my wallet.
>
> I ensure that web services are 100% functional for deprecated browsers,
> and there is lingering pressure-especially from the public wing of our
> institution (which I totally understand and, in the past, sympathized with)
> to present identical experiences across browsers. But you know what I did
> today? I sinned. From our global script, if modernizr detects that the
> browser is lt-ie9, it appends just below the navbar a subtle notice: "Did
> you know that your version of Internet Explorer is several years old? Why
> not give Firefox, Google Chrome, or Safari a try?"*
>
> In most circles this is considered the most heinous practice. But, you
> know, I can no longer passively stand by and see IE8 rank above the others
> when I give the analytics report to our web committee. Nope. The first step
> in this process was dropping all support for IE7 / Compatibility Mode a few
> months ago. Now that Google, jQuery, and others will soon drop support for
> IE8 - its time to politely join-in and make luddite patrons aware. IMHO,
> anyway.
>
> Already, old IE users get the raw end of the bargain because just viewing
> our website makes several additional server requests to pull additional CSS
> and JS bloat, not to mention all the images graphics they don't support.
> Thankfully, IE8 is cool with icon fonts, otherwise I'd be weeping at my
> desk.
>
> Now, why haven't I extended this behavior to browsers with limited support
> for, say, css gradients? That's trickier. A user might have the latest HTC
> phone but opt to surf in Opera Mini. There are too many variables and too
> many webkits (etc.). With old IE you can infer that a.) the user has a lap-
> or desktop, and [more importantly] b.) that old IE will never be a phone.
>
> Anyway,
>
> This is a really small-potatoes rant / action, but in a culture of all
> accessibility / never pressuring the user / whatever, it feels momentous. I
> kind of feel stupid getting all high and mighty about it. What do you think?
>
> Michael | Front End Librarian | www.ns4lib.com
>
> * Why, you may ask, did I not suggest IE9? Well, IE9 isn't exactly the
> experience we'd prefer them to have, but also according to our analytics
> the huge majority of old IE users are on Windows XP - where 9 isn't an
> option anyway. Eventually, down the road, we'll encourage IE9ers to upgrade
> too (once things like flexbox become standard), and at least they should
> have the option to try IE10.
>



-- 
Bill Dueber
Library Systems Programmer
University of Michigan Library