IE error can't debug!

[flyin...@gmail.com]

When i run the debugger, everything runs fine. however, when I try
using ie itself I sometimes get.

540 function
com_google_gwt_dom_client_Element_is__Lcom_google_gwt_core_client_JavaScriptObject_2
(o){
541 if (o.nodeType) {
542 return o.nodeType == 1;
543 }
544 return false;
545 }


Line:541
Char: 3
Error: Object required
Code: 0
URL: http://localhost:8080/instantquoteprogram/F7AE60F70A19094FD6CD8F6548AE3E8F.cache.html



This only happens in IE. Anyone have any clue what might cause this
error?

[flyin...@gmail.com]

Oh ya, after this error happens, nothing is clickable anymore.

On Apr 1, 9:48 am, "flyingb...@gmail.com" <flyingb...@gmail.com>
wrote:

> When i run the debugger, everything runs fine. however, when I try
> using ie itself I sometimes get.
>
> 540 function
> com_google_gwt_dom_client_Element_is__Lcom_google_gwt_core_client_JavaScriptObject_2
> (o){
> 541   if (o.nodeType) {
> 542     return o.nodeType == 1;
> 543   }
> 544   return false;
> 545 }
>
> Line:541
> Char: 3
> Error: Object required
> Code: 0

> URL:http://localhost:8080/instantquoteprogram/F7AE60F70A19094FD6CD8F6548A...

[flyin...@gmail.com]

When the script debugor editor runs it shows this:

function com_google_gwt_event_shared_HandlerManager_
$fireEvent__Lcom_google_gwt_event_shared_HandlerManager_2Lcom_google_gwt_event_shared_GwtEvent_2
(this$static, event){
var oldSource;
if (event.com_google_gwt_event_shared_GwtEvent_dead) {
event.revive__();
}
oldSource = event.com_google_gwt_event_shared_GwtEvent_source;
event.com_google_gwt_event_shared_GwtEvent_source = this
$static.com_google_gwt_event_shared_HandlerManager_source;
try {
++this
$static.com_google_gwt_event_shared_HandlerManager_firingDepth;
com_google_gwt_event_shared_HandlerManager$HandlerRegistry_
$fireEvent__Lcom_google_gwt_event_shared_HandlerManager
$HandlerRegistry_2Lcom_google_gwt_event_shared_GwtEvent_2Z(this
$static.com_google_gwt_event_shared_HandlerManager_registry, event,
this
$static.com_google_gwt_event_shared_HandlerManager_isReverseOrder);
}
finally {
--this
$static.com_google_gwt_event_shared_HandlerManager_firingDepth;
if (this
$static.com_google_gwt_event_shared_HandlerManager_firingDepth == 0) {
com_google_gwt_event_shared_HandlerManager_
$handleQueuedAddsAndRemoves__Lcom_google_gwt_event_shared_HandlerManager_2
(this$static);
}
}
if (oldSource == null) {
event.com_google_gwt_event_shared_GwtEvent_dead = true;
event.com_google_gwt_event_shared_GwtEvent_source = null;
}
else {
event.com_google_gwt_event_shared_GwtEvent_source = oldSource;
}
}



On Apr 1, 9:50 am, "flyingb...@gmail.com" <flyingb...@gmail.com>
wrote:

[Vitali Lovich]

Not sure what you mean - what debugger?  The proper way of debugging GWT is to run it in hosted mode, where it's your Java code that's being executed.

[flyin...@gmail.com]

But that does not show any errors at all. When I run the hosted mode
debugger no problems show up. It only happens when I am using IE. Does
not happen in the debug host mode thing. even though alert box says it
is internet explorer.

On Apr 1, 9:59 am, Vitali Lovich <vlov...@gmail.com> wrote:
> Not sure what you mean - what debugger?  The proper way of debugging GWT is
> to run it in hosted mode, where it's your Java code that's being executed.
>

> On Wed, Apr 1, 2009 at 12:55 PM, flyingb...@gmail.com
> <flyingb...@gmail.com>wrote:

[Vitali Lovich]

The hosted mode browser is some hacked up version of IE6 AFAIK.  I am going to say that
even if you compile in DETAILED mode, trying to track down a problem in the javascript code itself is pretty difficult - probably a waste of time.

Try putting logging statements in your code (either log to some textbox or to using RPC to a server) to try and trace the execution of your program.  Since it seems its a problem in the event handling, I'd suggesting putting those trace statements anywhere you have an event handler.

[flyin...@gmail.com]

I think my last msg didnt post. I think I figure out the problem. It
has to do with DialogBox.

I had old code that I wasn't using anymore. It should not do anything
though. I not sure why it is causing an error in the program. But it
is caused by the dialogbox i think

[meng]

just edit the compiled script, change

if (o.nodeType) {

to

if (o && o.nodeType) {

[Vitali Lovich]

That just introduces more problems.

1)  Not maintainable - you have to remember to change it every single time you compile.

2)  It's a hack.  Without understanding the real cause, you are just asking for more problems later on.  The problem could resurface elsewhere & you won't know why & you'll then have multiple places that need editing.
3) The problem is in your code - fix that & you won't have problems.  If it's a compiler issue (unlikely in this case), then file a bug & the GWT engineers will fix it for everyone.

PS - just to correct myself from before: on Windows, GWT uses whatever version of IE you have installed.

[Thomas Broyer]

On 28 mai, 14:57, Vitali Lovich <vlov...@gmail.com> wrote:
> That just introduces more problems.
> 1)  Not maintainable - you have to remember to change it every single time
> you compile.
> 2)  It's a hack.  Without understanding the real cause, you are just asking
> for more problems later on.  The problem could resurface elsewhere & you
> won't know why & you'll then have multiple places that need editing.
> 3) The problem is in your code - fix that & you won't have problems.  If
> it's a compiler issue (unlikely in this case), then file a bug & the GWT
> engineers will fix it for everyone.

This is a known issue (discussed several times here and/or on GWTC),
see also http://code.google.com/p/google-web-toolkit/issues/detail?id=3663

> PS - just to correct myself from before: on Windows, GWT uses whatever
> version of IE you have installed.

...and keep in mind that the IE8 ActiveX component runs in
"compatibility mode" by default, so the HostedMode/GWTShell will be in
"compatibility mode".

[jorge]

Hi all,

As Thomas says, it's a known bug for which there has not been solution
for a long time (I had the same problem). The hack is simple but as
the GWT team doesn't know why this is happening (it shouldn't) they
have not solved it yet.

The bug doesn't appear in firefox, only in ie, but it's true that
modifying the javascript is not maintainable.

The workaround I have used is also a hack but a little more
maintenable:
- create a package on your project: com.google.gwt.user.client.ui
- copy there the source code of DialogBox and PopupPanel from GWT
- remame them to DialogBox2 and PopupPanel2 or similar (update all
internal references in the files, I can send you the modified sources
if you want)
- do the JS hack in those files
- use those classes instead of GWT's on your project.

You can use those classes while you wait for the bug to be solved.

Hope it helps you,

Bye

On May 29, 1:06 am, Thomas Broyer <t.bro...@gmail.com> wrote:
> On 28 mai, 14:57, Vitali Lovich <vlov...@gmail.com> wrote:
>
> > That just introduces more problems.
> > 1)  Not maintainable - you have to remember to change it every single time
> > you compile.
> > 2)  It's a hack.  Without understanding the real cause, you are just asking
> > for more problems later on.  The problem could resurface elsewhere & you
> > won't know why & you'll then have multiple places that need editing.
> > 3) The problem is in your code - fix that & you won't have problems.  If
> > it's a compiler issue (unlikely in this case), then file a bug & the GWT
> > engineers will fix it for everyone.
>
> This is a known issue (discussed several times here and/or on GWTC),

> see alsohttp://code.google.com/p/google-web-toolkit/issues/detail?id=3663

[Thomas Broyer]

On May 29, 11:26 am, jorge <jorgelopezma...@googlemail.com> wrote:
> Hi all,
>
> As Thomas says, it's a known bug for which there has not been solution
> for a long time (I had the same problem). The hack is simple but as
> the GWT team doesn't know why this is happening (it shouldn't) they
> have not solved it yet.

It is actually fixed, but the bug has been marked as Fixed instead of
FixedNotReleased (that's why I didn't find it earlier, but this time I
started from the source code and SVN commit logs):
http://code.google.com/p/google-web-toolkit/issues/detail?id=3644

So I believe issue 3663 should be marked FixedNotReleased too (and/or
Duplicate and merged to issue 3644):
http://code.google.com/p/google-web-toolkit/issues/detail?id=3663