Presumably the fact that it can have trouble correctly displaying sites coded exclusively for Microsoft's bastardised proprietary non-standard HTML coding.
Yes indeed, it needs its own different bastardised proprietary non-standard [] coding.
I left out HTML in the above sentence because that's not really where I've run into problems - it's much more of an issue with
JS and CSS.
Wittering on about how many standards it obeys doesn't hack it I'm afraid ... the browser is actually a very crude and immature platform for developing applications (as opposed to web sites), and large numbers of blindingly obvious things that you're bound to want to do
have no standard way of doing them, so you have no choice but to find out which non-standard trick works for each browser - for a practical application
you cannot code to the standards because the standards simply have nothing to say about lots of features that you actually need.
So, code exclusively for IE and you get a much better experience ... so long as you are using IE. Code exclusively for Firefox and it won't work on IE. Carefully code to make sure that what you're doing works fine on both IE and FF ... and it doesn't work on Opera.
By the way you can't actually tell just from looking at the code whether someone has put in a terrible hack for one browser or another. For example here is one of my scenarios:
(1) I start out by trying to write standard code, and only deviate from that when the standard code doesn't work.
(2) For this particular feature the standard code worked on IE but not on FF.
(3) So I fiddled around and found a non-standard bodge that worked on FF.
(4) But this didn't work on IE ... however I was then able to work around IE's correct handling of this wrong code by putting an IE-specific section of code in.
So anyone reading the code would discover some horrible IE specific kludge code, but it would not be obvious (unless they troubled to read the comments) that I'd put it there to work around a failure of FF to implement the standards.
===========
Here's a good Firefox nasty. Have a form on your web page, with a button on it, and no action assigned to either the form or the button ('cos you haven't written that bit yet). Clicking on the button should do nothing, yes? That's indeed what happens in IE. But in Firefox it doesn't quite do nothing - it crashes out any pending AJAX calls with obscure and undocumented error messages, which if you search for lead you straight into the Firefox bug database, with no sign that anyone is remotely interested in fixing it.
Workaround? Remove the form. So now you have a disembodied free button that is illegally, according to the standards, not in any form. This will fail all the validation tests recommended by the "everything must be standard" weenies, but
it works on all the browsers I've tried.