jQuery Forum

Hi. I was hoping someone could shed some light on a problem I'm
seeing w/UI 1.7.2 that isn't present w/UI 1.5.3.
I've got some mildly complicated code (or I'd post the whole thing
here) that relies on draggable and droppable. It works fine in
1.2.6/1.5.3, but throws an error under 1.3.2/1.7.2 inside of the
following code (which is part of jQuery UI):
$.ui.plugin.add("draggable", "cursor", {
start: function(event, ui) {
var t = $('body'), o = $(this).data('draggable').options;
if (t.css("cursor")) o._cursor = t.css("cursor");
t.css("cursor", o.cursor);
},
stop: function(event, ui) {
var o = $(this).data('draggable').options;
<===== $(this).data("draggable") is undefined
if (o._cursor) $('body').css("cursor", o._cursor);
}
});
The code in 1.7.2 apparently expects that the draggable element has
some associated data but it seems not to be there. Again, this isn't
a problem in 1.5.3.
The logic in question is using draggable and droppable to implement
dragging grid (table) columns around to reorder the columns. When a
column is dropped, it blows away the elements in the table's thead and
tbody and
regenerates the HTML for them using the new column sequence.
Stepping through the code, I can see that my logic that dumps and re-
renders the HTML occurs before the above problem line is hit. Which I
guess might be the problem. But then why isn't it a problem in 1.5.3?
I'm kind of stumbling around a bit on this so any thoughts as to what
might be the cause/solution for this would be very much appreciated.
Thanks,
Andy

Couple questions for the jQuery internals-aware folks. This is a bit
long, sorry for that.
I'm working on beefing up error handling in a web application. I want
to ensure that I'm catching all exceptions and handling them via my
own UI system. The window.onerror function is part of the solution,
but there's a known problem with Firefox (https://bugzilla.mozilla.org/
show_bug.cgi?id=312448) that prevents onerror being called when the
exception is thrown inside an event handler.
In looking for a solution that didn't require mucking about with each
instance of event binding logic throughout the app, I ran across what
seems like a reasonable solution here: (http://blogs.cozi.com/tech/
2008/04/javascript-error-tracking-why-windowonerror-is-not-
enough.html). The general idea is to override bind() and replace the
passed in handler function with a function that wraps the original
handler function in a try-catch block.
So the code would look something like this:
var jqBindFn = jQuery.fn.bind;
$.fn.bind = function(type, data, fn) {
if (!fn && data && typeof data === 'function') {
fn = data;
data = null;
}
if (fn) {
var ofn = fn;
var wfn = function() {
try {
ofn.apply(this, arguments);
}
catch ( ex ) {
someGlobalErrHandler( ex );
}
};
fn = wfn;
}
return jqBindFn.call(this, type, data, fn);
};
Since all other event binding goes through bind(), it seems like this
will easily add exception handling to all event handlers in my app.
So I have two questions about this:
(1) I haven't been using Javascript terribly long, so I'm not really
sure if this approach has any significant costs or downsides. I mean,
there's a performance hit by having the extra function calls. And
again I suppose for the exception handling. But are there potential
memory leak issues or is the performance hit likely to increase the
longer a user uses an app with that "fix" in it?
(2) It sort of seems like Mozilla isn't in any giant hurry to fix
there bug. It probably doesn't make sense to add code to jQuery core
to work around such an obscure bug for it's own sake. But I wonder if
adding some exception handling capability to jQuery core makes any
sense? It could do what it does for other areas of Javascript which
is mitigate the browser differences. To be complete, it would need to
employ onerror, add a try-catch block to document.ready(), and do
something similar to the above for event handlers (only in mozilla).
Thoughts? Is this just stupid?

I'm working on a small plugin that extends a checkbox to behave as if
it has 3 states. Basically it just adds an anchor above the
underlying checkbox, intercepts mouse clicks, and cycles the checkbox
between:
disabled + unchecked
enabled + unchecked
enabled + checked
It works fine in FF. But in IE7, for some reason, the click events
are always sent to the underlying checkbox instead of the overlaid
anchor. The anchor is sized to completely overlay the checkbox (and
in fact, I made it's slightly bigger so clicking at the very edges
works).
Incidentally, I'm using jQuery 1.2.6 since this is part of a much
larger web application and we haven't yet worked out the
incompatibilities we're seeing with the latest jQuery version.
Here's the plugin code (the alerts are for debugging in IE; please
don't beat up on me too bad - it's an early work-in-progress):
(function($) {
$.fn.checkbox3s = function(options) {
var opts = $.extend({}, $.fn.defaults, options);
return this.each(function() {
var proxy, cbx = $(this), parent = cbx.parent();
parent.css({position:"relative"});
proxy = $("<a href='#'></a>").css({
position:"absolute",
display:"block",
padding:"0", margin:"0",
textDecoration:"none",
left: "0px", top: "0px",
width: "100%", height: "100%"
}).appendTo(parent).click(function(e) {
if (cbx.is(":disabled")) {
alert("disabled - enabling and unchecking");
cbx.attr({disabled: false, checked:false});
} else if (cbx.is(":checked")){
alert("enabled & checked - unchecking & disabling");
cbx.attr({checked:false, disabled:true});
} else {
alert("enabled & unchecked - check");
cbx.attr({checked:true});
}
return false;
});
cbx.attr({checked:false, disabled:true}).css
({position:"relative"/*, zIndex:499*/}).click(function(){return
false;});
});
};
$.fn.checkbox3s.defaults = {};
})(jQuery);
Anyone know why this would work okay in FF but not in IE (besides the
obviousness that IE sucks)?
Thanks,
Andy

Any word on the much-anticipated jQuery conference 2009?

powers-that-be:
It's a good thing that this group is moderated.
But I was wondering how many messages I need to post before I get
unmoderated? It's kind of a drag to respond to a thread and find that
by the time the response becomes visible, the response has become
obsolete.
Also, does being unmoderated in one jQuery discussion group,
unmoderate me in the others? Or are they individual? I'm also a
member of the jQuery UI group.
Thanks,
Andy

If I have a table:
<table>
<tbody>
<tr><td><span></span></td>/*many more cells here*/</tr>
<tr><td><span></span></td>/*many more cells here*/</tr>
// many more rows here...
<tbody>
</table>
And I'm handed a reference to a <span> inside one of the table cells,
what's the quickest method to obtain the row that contains the span?
Is this as fast as it gets?
$(theSpan).parents('tr:first');
Does that even work? Does ":first" work when climbing upwards through
the ancestry?
Thanks,
Andy