disconnected nodes and css

gistfile1.js

// suppose we have this in css
// .hidden { display: none; }
// .inline { display: inline; }

// then in javascrpit
$('<div class="hidden"/>').css("display");
// for FF – "none"
// for IE – "block"
// For Chrome, Opera – ""

$('<div class="inline"/>').css("display");
// for FF – "inline"
// for IE – "block"
// For Chrome, Opera – ""

// with current code if you do
$('<div class="hidden"/>').fadeIn() | show( 0 ) | fadeTo();
$('<div class="inline"/>').fadeIn() | show( 0 ) | fadeTo();
// it will set display to "block"  

// but with pull 774
$('<div class="hidden"/>').fadeIn() | show( 0 ) | fadeTo();
// FF will change display, when others will not

$('<div class="inline"/>').fadeIn() | show( 0 ) | fadeTo();
// display will not be set for all browsers

// but for
$('<div class="hidden"/>').show();
$('<div class="inline"/>').show();
// display will always be set

// so we should also check for disconnected nodes, something like this – 
jQuery.css( elem, "display" ) === "none" || !jQuery.contains( elem.ownerDocument.documentElement, elem );

// Also, there issue with height: 0; elements
// now it's like this – http://jsfiddle.net/z9r2e/
// because test for visibility in showHide and defaultPrefilter functions are different

// but with 774 change test will be identical, so it will always hide elements like that
// really, i don't see a problem with it, for me, it looks like more righteous behavior,
// for me, present behavior looks like bug

// and this
jQuery.css( elem, "display" ) === "none" || !jQuery.contains( elem.ownerDocument.documentElement, elem );

// still faster than this
jQuery( elem ).is(":hidden");
// http://jsperf.com/ishidden-vs-hidden