jQuery Forum

You say that you still have problems if you split apart the query.
So in this case $button.append( "[" ) fails - correct?
What happens if you do:
$button.append( document.createTextNode("[") )
--John

Ok, narrowed it down.
Someone tells me that Konqueror isn't working for that page either so it
looks like IE8 and Konqueror may share this behavior.
From the looks of it $button.append( "[" ); is working.
$('#bodyContent').append( "[" ); // works
$('#bodyContent').append( "[", "]" ); // "Invalid argument" in IE8 console, works in FF Firebug console
$('#bodyContent').append( "asdf", "asdf" ); // "Invalid argument" in IE8 console, works in FF Firebug console
$('#bodyContent').append( document.createTextNode("["), "]" ); // works
$('#bodyContent').append( "<span>foo</span>", "]" ); // in IE8 console only inserts the ']' the foo span is not inserted, works in FF Firebug console
$('#bodyContent').append( "<span>asdf</span>", "<span>qwerty</span>" ); // in IE8 console only inserts the qwerty span the asdf span is not inserted, works in FF Firebug console
So this looks like an IE8 (maybe Konqueror) issue where when .domManip
is used (by something like .append) with more than one argument with the
first being a string a "Invalid argument" error is thrown if the first
argument is not a valid node string. If the first argument is a valid
node string, instead the first argument is ignored and the rest are
inserted.
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
John Resig wrote:
> You say that you still have problems if you split apart the query.
>
> So in this case $button.append( "[" ) fails - correct?
>
> What happens if you do:
>
> $button.append( document.createTextNode("[") )
>
> --John
>
>
>
> On Mon, Jun 22, 2009 at 2:09 PM, Daniel
> Friesen<nadir.seen.fire@gmail.com>

This is very helpful analysis. I've added it to my todo list. Could
you file a ticket with your test cases, as well? Thanks!
http://dev.jquery.com/newticket
--John

http://dev.jquery.com/ticket/4805
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
John Resig wrote:
> This is very helpful analysis. I've added it to my todo list. Could
> you file a ticket with your test cases, as well? Thanks!
> http://dev.jquery.com/newticket
>
> --John
>
>
>
> On Mon, Jun 22, 2009 at 4:33 PM, Daniel
> Friesen<nadir.seen.fire@gmail.com> wrote:
>
>> Ok, narrowed it down.
>> Someone tells me that Konqueror isn't working for that page either so it
>> looks like IE8 and Konqueror may share this behavior.
>> From the looks of it $button.append( "[" ); is working.
>>
>> $('#bodyContent').append( "[" ); // works
>> $('#bodyContent').append( "[", "]" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>> $('#bodyContent').append( "asdf", "asdf" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>> $('#bodyContent').append( document.createTextNode("["), "]" ); // works
>> $('#bodyContent').append( "<span>foo</span>", "]" ); // in IE8 console only inserts the ']' the foo span is not inserted, works in FF Firebug console
>> $('#bodyContent').append( "<span>asdf</span>", "<span>qwerty</span>" ); // in IE8 console only inserts the qwerty span the asdf span is not inserted, works in FF Firebug console
>>
>> So this looks like an IE8 (maybe Konqueror) issue where when .domManip
>> is used (by something like .append) with more than one argument with the
>> first being a string a "Invalid argument" error is thrown if the first
>> argument is not a valid node string. If the first argument is a valid
>> node string, instead the first argument is ignored and the rest are
>> inserted.
>>
>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>
>> John Resig wrote:
>>
>>> You say that you still have problems if you split apart the query.
>>>
>>> So in this case $button.append( "[" ) fails - correct?
>>>
>>> What happens if you do:
>>>
>>> $button.append( document.createTextNode("[") )
>>>
>>> --John
>>>
>>>
>>>
>>> On Mon, Jun 22, 2009 at 2:09 PM, Daniel
>>> Friesen<nadir.seen.fire@gmail.com>

I ran into another IE issue with unclosed tag syntax inside node
creation (I personally like to leave the / out unless necessary most of
the time, especially since it flows fairly nice with .text(userVar)).
http://jsbin.com/ejote
<span></span>
<span />
<span/>
<span>
<span class=foo />
All work in ie7, ie8, opera, firefox, and midori.
However:
<span class=foo>
<span class="foo">
Fail in ie7 and ie8 while they work in firefox, midori, and opera.
Filed as: http://dev.jquery.com/ticket/4806
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
John Resig wrote:
> This is very helpful analysis. I've added it to my todo list. Could
> you file a ticket with your test cases, as well? Thanks!
> http://dev.jquery.com/newticket
>
> --John
>
>
>
> On Mon, Jun 22, 2009 at 4:33 PM, Daniel
> Friesen<nadir.seen.fire@gmail.com> wrote:
>
>> Ok, narrowed it down.
>> Someone tells me that Konqueror isn't working for that page either so it
>> looks like IE8 and Konqueror may share this behavior.
>> From the looks of it $button.append( "[" ); is working.
>>
>> $('#bodyContent').append( "[" ); // works
>> $('#bodyContent').append( "[", "]" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>> $('#bodyContent').append( "asdf", "asdf" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>> $('#bodyContent').append( document.createTextNode("["), "]" ); // works
>> $('#bodyContent').append( "<span>foo</span>", "]" ); // in IE8 console only inserts the ']' the foo span is not inserted, works in FF Firebug console
>> $('#bodyContent').append( "<span>asdf</span>", "<span>qwerty</span>" ); // in IE8 console only inserts the qwerty span the asdf span is not inserted, works in FF Firebug console
>>
>> So this looks like an IE8 (maybe Konqueror) issue where when .domManip
>> is used (by something like .append) with more than one argument with the
>> first being a string a "Invalid argument" error is thrown if the first
>> argument is not a valid node string. If the first argument is a valid
>> node string, instead the first argument is ignored and the rest are
>> inserted.
>>
>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>
>> John Resig wrote:
>>
>>> You say that you still have problems if you split apart the query.
>>>
>>> So in this case $button.append( "[" ) fails - correct?
>>>
>>> What happens if you do:
>>>
>>> $button.append( document.createTextNode("[") )
>>>
>>> --John
>>>
>>>
>>>
>>> On Mon, Jun 22, 2009 at 2:09 PM, Daniel
>>> Friesen<nadir.seen.fire@gmail.com>

Unfortunately, there's not really a whole lot that we can do on our
end to fix a problem like that - short of including a full HTML
parser/DOM generator.
Now, I've written one:
http://ejohn.org/blog/pure-javascript-html-parser/
but it's probably just better to make sure that your injected HTML is
well-formed.
--John

Ok, I agree that $('<div><span></span>') is just plain malformed html
and doesn't need support.
But to be honest $('<span>') looks like we're saying to create a span
node, and likewise $('<span attr=val>') like we're trying to create a
span with a single attr.
Originally when I found out about $('<span>') I wasn't thinking HTML, I
originally thought "Oh yay, jQuery has a shortcut for
document.createElement, and after when I saw that $('<span foo=bar />')
was possible I didn't think HTML, I still thought, createElement + attr
setting.
Would a patch that just makes $('<span attr=val>') work properly across
browsers be accepted?
That should be something simple like checking if it starts with < and
ends with >, contains no other <>'s and doesn't have a / before the >
then insert a single / before the last > before continuing.
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
John Resig wrote:
> Unfortunately, there's not really a whole lot that we can do on our
> end to fix a problem like that - short of including a full HTML
> parser/DOM generator.
>
> Now, I've written one:
> http://ejohn.org/blog/pure-javascript-html-parser/
>
> but it's probably just better to make sure that your injected HTML is
> well-formed.
>
> --John
>
>
>
> On Mon, Jun 22, 2009 at 6:45 PM, Daniel
> Friesen<nadir.seen.fire@gmail.com> wrote:
>
>> I ran into another IE issue with unclosed tag syntax inside node
>> creation (I personally like to leave the / out unless necessary most of
>> the time, especially since it flows fairly nice with .text(userVar)).
>>
>> http://jsbin.com/ejote
>>
>> <span></span>
>> <span />
>> <span/>
>> <span>
>> <span class=foo />
>>
>> All work in ie7, ie8, opera, firefox, and midori.
>>
>> However:
>> <span class=foo>
>> <span class="foo">
>>
>> Fail in ie7 and ie8 while they work in firefox, midori, and opera.
>>
>> Filed as: http://dev.jquery.com/ticket/4806
>>
>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>
>> John Resig wrote:
>>
>>> This is very helpful analysis. I've added it to my todo list. Could
>>> you file a ticket with your test cases, as well? Thanks!
>>> http://dev.jquery.com/newticket
>>>
>>> --John
>>>
>>>
>>>
>>> On Mon, Jun 22, 2009 at 4:33 PM, Daniel
>>> Friesen<nadir.seen.fire@gmail.com> wrote:
>>>
>>>
>>>> Ok, narrowed it down.
>>>> Someone tells me that Konqueror isn't working for that page either so it
>>>> looks like IE8 and Konqueror may share this behavior.
>>>> From the looks of it $button.append( "[" ); is working.
>>>>
>>>> $('#bodyContent').append( "[" ); // works
>>>> $('#bodyContent').append( "[", "]" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>>>> $('#bodyContent').append( "asdf", "asdf" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>>>> $('#bodyContent').append( document.createTextNode("["), "]" ); // works
>>>> $('#bodyContent').append( "<span>foo</span>", "]" ); // in IE8 console only inserts the ']' the foo span is not inserted, works in FF Firebug console
>>>> $('#bodyContent').append( "<span>asdf</span>", "<span>qwerty</span>" ); // in IE8 console only inserts the qwerty span the asdf span is not inserted, works in FF Firebug console
>>>>
>>>> So this looks like an IE8 (maybe Konqueror) issue where when .domManip
>>>> is used (by something like .append) with more than one argument with the
>>>> first being a string a "Invalid argument" error is thrown if the first
>>>> argument is not a valid node string. If the first argument is a valid
>>>> node string, instead the first argument is ignored and the rest are
>>>> inserted.
>>>>
>>>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>>>
>>>> John Resig wrote:
>>>>
>>>>
>>>>> You say that you still have problems if you split apart the query.
>>>>>
>>>>> So in this case $button.append( "[" ) fails - correct?
>>>>>
>>>>> What happens if you do:
>>>>>
>>>>> $button.append( document.createTextNode("[") )
>>>>>
>>>>> --John
>>>>>
>>>>>
>>>>>
>>>>> On Mon, Jun 22, 2009 at 2:09 PM, Daniel
>>>>> Friesen<nadir.seen.fire@gmail.com>

Why support bad mark-up? Adding the slash costs nothing.

I'm hesitant to support that since it tends to promote passing in
malformed (X)HTML. At least with <span attr='val'/> it's obvious that
it should be a standalone.
--John

http://www.google.com/codesearch?hl=en&lr=&q=\%24\([%27%22]%3C[^%3C%3E%25]%2B[^%2F]%3E[%27%22]\)+lang%3Ajs&sbtn=Search
I would have thought there'd be more. Frankly it is a sane way to think,
jquery is dom manipulation and $('<td>') looks a lot like tag creation
rather than something than html being mixed into jQuery.
I for one have been thinking of $('<...>') as tag creation rather than
raw html, or thinking that jQuery did it's own quick parsing for ages.
Now I'll have to go back through 6 months of code to track down all the
code using that pattern.
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
John Resig wrote:
> I'm hesitant to support that since it tends to promote passing in
> malformed (X)HTML. At least with <span attr='val'/> it's obvious that
> it should be a standalone.
>
> --John
>
>
>
> On Tue, Jun 23, 2009 at 12:27 AM, Daniel
> Friesen<nadir.seen.fire@gmail.com> wrote:
>
>> Ok, I agree that $('<div><span></span>') is just plain malformed html
>> and doesn't need support.
>> But to be honest $('<span>') looks like we're saying to create a span
>> node, and likewise $('<span attr=val>') like we're trying to create a
>> span with a single attr.
>> Originally when I found out about $('<span>') I wasn't thinking HTML, I
>> originally thought "Oh yay, jQuery has a shortcut for
>> document.createElement, and after when I saw that $('<span foo=bar />')
>> was possible I didn't think HTML, I still thought, createElement + attr
>> setting.
>>
>> Would a patch that just makes $('<span attr=val>') work properly across
>> browsers be accepted?
>>
>> That should be something simple like checking if it starts with < and
>> ends with >, contains no other <>'s and doesn't have a / before the >
>> then insert a single / before the last > before continuing.
>>
>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>
>> John Resig wrote:
>>
>>> Unfortunately, there's not really a whole lot that we can do on our
>>> end to fix a problem like that - short of including a full HTML
>>> parser/DOM generator.
>>>
>>> Now, I've written one:
>>> http://ejohn.org/blog/pure-javascript-html-parser/
>>>
>>> but it's probably just better to make sure that your injected HTML is
>>> well-formed.
>>>
>>> --John
>>>
>>>
>>>
>>> On Mon, Jun 22, 2009 at 6:45 PM, Daniel
>>> Friesen<nadir.seen.fire@gmail.com> wrote:
>>>
>>>
>>>> I ran into another IE issue with unclosed tag syntax inside node
>>>> creation (I personally like to leave the / out unless necessary most of
>>>> the time, especially since it flows fairly nice with .text(userVar)).
>>>>
>>>> http://jsbin.com/ejote
>>>>
>>>> <span></span>
>>>> <span />
>>>> <span/>
>>>> <span>
>>>> <span class=foo />
>>>>
>>>> All work in ie7, ie8, opera, firefox, and midori.
>>>>
>>>> However:
>>>> <span class=foo>
>>>> <span class="foo">
>>>>
>>>> Fail in ie7 and ie8 while they work in firefox, midori, and opera.
>>>>
>>>> Filed as: http://dev.jquery.com/ticket/4806
>>>>
>>>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>>>
>>>> John Resig wrote:
>>>>
>>>>
>>>>> This is very helpful analysis. I've added it to my todo list. Could
>>>>> you file a ticket with your test cases, as well? Thanks!
>>>>> http://dev.jquery.com/newticket
>>>>>
>>>>> --John
>>>>>
>>>>>
>>>>>
>>>>> On Mon, Jun 22, 2009 at 4:33 PM, Daniel
>>>>> Friesen<nadir.seen.fire@gmail.com> wrote:
>>>>>
>>>>>
>>>>>
>>>>>> Ok, narrowed it down.
>>>>>> Someone tells me that Konqueror isn't working for that page either so it
>>>>>> looks like IE8 and Konqueror may share this behavior.
>>>>>> From the looks of it $button.append( "[" ); is working.
>>>>>>
>>>>>> $('#bodyContent').append( "[" ); // works
>>>>>> $('#bodyContent').append( "[", "]" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>>>>>> $('#bodyContent').append( "asdf", "asdf" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>>>>>> $('#bodyContent').append( document.createTextNode("["), "]" ); // works
>>>>>> $('#bodyContent').append( "<span>foo</span>", "]" ); // in IE8 console only inserts the ']' the foo span is not inserted, works in FF Firebug console
>>>>>> $('#bodyContent').append( "<span>asdf</span>", "<span>qwerty</span>" ); // in IE8 console only inserts the qwerty span the asdf span is not inserted, works in FF Firebug console
>>>>>>
>>>>>> So this looks like an IE8 (maybe Konqueror) issue where when .domManip
>>>>>> is used (by something like .append) with more than one argument with the
>>>>>> first being a string a "Invalid argument" error is thrown if the first
>>>>>> argument is not a valid node string. If the first argument is a valid
>>>>>> node string, instead the first argument is ignored and the rest are
>>>>>> inserted.
>>>>>>
>>>>>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>>>>>
>>>>>> John Resig wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>> You say that you still have problems if you split apart the query.
>>>>>>>
>>>>>>> So in this case $button.append( "[" ) fails - correct?
>>>>>>>
>>>>>>> What happens if you do:
>>>>>>>
>>>>>>> $button.append( document.createTextNode("[") )
>>>>>>>
>>>>>>> --John
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Jun 22, 2009 at 2:09 PM, Daniel
>>>>>>> Friesen<nadir.seen.fire@gmail.com>

To be clear: jQuery supports the pattern, it just requires the closing /.
$("<td/>") // ok
$("<td>") // ambiguous, malformed, not ok
$("<a href='http://google.com/'/>") // ok
$("<a href='http://google.com/'>") // ambiguous, malformed, not ok
--John

On Tue, 23 Jun 2009 12:40:11 -0400, John Resig <jeresig@gmail.com>

Yeah, I think all outstanding bugs with that were fixed in 1.3.0.
--John

Can we update the documentation to also note that unclosed tags aren't
supported.
Malformed html is obvious, but people don't always think of a single tag
as malformed html.
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
John Resig wrote:
> To be clear: jQuery supports the pattern, it just requires the closing /.
>
> $("<td/>") // ok
> $("<td>") // ambiguous, malformed, not ok
>
> $("<a href='http://google.com/'/>") // ok
> $("<a href='http://google.com/'>") // ambiguous, malformed, not ok
>
> --John
>
>
>
> On Tue, Jun 23, 2009 at 10:37 AM, Daniel
> Friesen<nadir.seen.fire@gmail.com> wrote:
>
>> http://www.google.com/codesearch?hl=en&lr=&q=\%24\([%27%22]%3C[^%3C%3E%25]%2B[^%2F]%3E[%27%22]\)+lang%3Ajs&sbtn=Search
>>
>> I would have thought there'd be more. Frankly it is a sane way to think,
>> jquery is dom manipulation and $('<td>') looks a lot like tag creation
>> rather than something than html being mixed into jQuery.
>> I for one have been thinking of $('<...>') as tag creation rather than
>> raw html, or thinking that jQuery did it's own quick parsing for ages.
>> Now I'll have to go back through 6 months of code to track down all the
>> code using that pattern.
>>
>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>
>> John Resig wrote:
>>
>>> I'm hesitant to support that since it tends to promote passing in
>>> malformed (X)HTML. At least with <span attr='val'/> it's obvious that
>>> it should be a standalone.
>>>
>>> --John
>>>
>>>
>>>
>>> On Tue, Jun 23, 2009 at 12:27 AM, Daniel
>>> Friesen<nadir.seen.fire@gmail.com> wrote:
>>>
>>>
>>>> Ok, I agree that $('<div><span></span>') is just plain malformed html
>>>> and doesn't need support.
>>>> But to be honest $('<span>') looks like we're saying to create a span
>>>> node, and likewise $('<span attr=val>') like we're trying to create a
>>>> span with a single attr.
>>>> Originally when I found out about $('<span>') I wasn't thinking HTML, I
>>>> originally thought "Oh yay, jQuery has a shortcut for
>>>> document.createElement, and after when I saw that $('<span foo=bar />')
>>>> was possible I didn't think HTML, I still thought, createElement + attr
>>>> setting.
>>>>
>>>> Would a patch that just makes $('<span attr=val>') work properly across
>>>> browsers be accepted?
>>>>
>>>> That should be something simple like checking if it starts with < and
>>>> ends with >, contains no other <>'s and doesn't have a / before the >
>>>> then insert a single / before the last > before continuing.
>>>>
>>>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>>>
>>>> John Resig wrote:
>>>>
>>>>
>>>>> Unfortunately, there's not really a whole lot that we can do on our
>>>>> end to fix a problem like that - short of including a full HTML
>>>>> parser/DOM generator.
>>>>>
>>>>> Now, I've written one:
>>>>> http://ejohn.org/blog/pure-javascript-html-parser/
>>>>>
>>>>> but it's probably just better to make sure that your injected HTML is
>>>>> well-formed.
>>>>>
>>>>> --John
>>>>>
>>>>>
>>>>>
>>>>> On Mon, Jun 22, 2009 at 6:45 PM, Daniel
>>>>> Friesen<nadir.seen.fire@gmail.com> wrote:
>>>>>
>>>>>
>>>>>
>>>>>> I ran into another IE issue with unclosed tag syntax inside node
>>>>>> creation (I personally like to leave the / out unless necessary most of
>>>>>> the time, especially since it flows fairly nice with .text(userVar)).
>>>>>>
>>>>>> http://jsbin.com/ejote
>>>>>>
>>>>>> <span></span>
>>>>>> <span />
>>>>>> <span/>
>>>>>> <span>
>>>>>> <span class=foo />
>>>>>>
>>>>>> All work in ie7, ie8, opera, firefox, and midori.
>>>>>>
>>>>>> However:
>>>>>> <span class=foo>
>>>>>> <span class="foo">
>>>>>>
>>>>>> Fail in ie7 and ie8 while they work in firefox, midori, and opera.
>>>>>>
>>>>>> Filed as: http://dev.jquery.com/ticket/4806
>>>>>>
>>>>>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>>>>>
>>>>>> John Resig wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>> This is very helpful analysis. I've added it to my todo list. Could
>>>>>>> you file a ticket with your test cases, as well? Thanks!
>>>>>>> http://dev.jquery.com/newticket
>>>>>>>
>>>>>>> --John
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Jun 22, 2009 at 4:33 PM, Daniel
>>>>>>> Friesen<nadir.seen.fire@gmail.com> wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> Ok, narrowed it down.
>>>>>>>> Someone tells me that Konqueror isn't working for that page either so it
>>>>>>>> looks like IE8 and Konqueror may share this behavior.
>>>>>>>> From the looks of it $button.append( "[" ); is working.
>>>>>>>>
>>>>>>>> $('#bodyContent').append( "[" ); // works
>>>>>>>> $('#bodyContent').append( "[", "]" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>>>>>>>> $('#bodyContent').append( "asdf", "asdf" ); // "Invalid argument" in IE8 console, works in FF Firebug console
>>>>>>>> $('#bodyContent').append( document.createTextNode("["), "]" ); // works
>>>>>>>> $('#bodyContent').append( "<span>foo</span>", "]" ); // in IE8 console only inserts the ']' the foo span is not inserted, works in FF Firebug console
>>>>>>>> $('#bodyContent').append( "<span>asdf</span>", "<span>qwerty</span>" ); // in IE8 console only inserts the qwerty span the asdf span is not inserted, works in FF Firebug console
>>>>>>>>
>>>>>>>> So this looks like an IE8 (maybe Konqueror) issue where when .domManip
>>>>>>>> is used (by something like .append) with more than one argument with the
>>>>>>>> first being a string a "Invalid argument" error is thrown if the first
>>>>>>>> argument is not a valid node string. If the first argument is a valid
>>>>>>>> node string, instead the first argument is ignored and the rest are
>>>>>>>> inserted.
>>>>>>>>
>>>>>>>> ~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]
>>>>>>>>
>>>>>>>> John Resig wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>> You say that you still have problems if you split apart the query.
>>>>>>>>>
>>>>>>>>> So in this case $button.append( "[" ) fails - correct?
>>>>>>>>>
>>>>>>>>> What happens if you do:
>>>>>>>>>
>>>>>>>>> $button.append( document.createTextNode("[") )
>>>>>>>>>
>>>>>>>>> --John
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Mon, Jun 22, 2009 at 2:09 PM, Daniel
>>>>>>>>> Friesen<nadir.seen.fire@gmail.com>

Ok, done:
http://docs.jquery.com/Core/jQuery#htmlownerDocument
--John

this is just a workaround i have used for this prepend/append issue in ie :