Announcement

**anstaafl** · 1 May 2013, 08:51

I'm still getting the crash with today's build.
Can you be more specific about running in compiled mode?

I'm running this in dev mode in eclipse Juno.
I updated the libraries, recompiled the source and flushed the browser cache.

I brought up the SGWT developer console.

SmartClient Version: SNAPSHOT_v9.0d_2013-05-01/LGPL Development Only (built 2013-05-01)

The funny thing is, the crash doesn't get displayed in the Log messages tab...

Code:

12:32:22.217:INFO:Log:initialized
12:32:23.826:WARN:Log:New Class ID: 'DataView' collides with ID of existing object with value 'function DataView() {
    [native code]
}'.  Existing object will be replaced.
This conflict would be avoided by disabling ISC Simple Names mode.  See documentation for further information.
12:32:24.031:INFO:Log:isc.Page is loaded

Yet FireFox says there's a crash and the eclipse development mode tab spewed the following.

Code:

12:33:16.513 [ERROR] [RestCacheAllData] Uncaught exception escaped

com.google.gwt.core.client.JavaScriptException: (null) @com.smartgwt.client.data.DataSource::getClientOnlyResponse(Lcom/smartgwt/client/data/DSRequest;[Lcom/smartgwt/client/data/Record;)([Java object: com.smartgwt.client.data.DSRequest@33355127, Java object: [Lcom.smartgwt.client.data.Record;@25392310]): null
    at com.google.gwt.dev.shell.BrowserChannelServer.invokeJavascript(BrowserChannelServer.java:249)
    at com.google.gwt.dev.shell.ModuleSpaceOOPHM.doInvoke(ModuleSpaceOOPHM.java:136)
    at com.google.gwt.dev.shell.ModuleSpace.invokeNative(ModuleSpace.java:571)
    at com.google.gwt.dev.shell.ModuleSpace.invokeNativeObject(ModuleSpace.java:279)
    at com.google.gwt.dev.shell.JavaScriptHost.invokeNativeObject(JavaScriptHost.java:91)
    at com.smartgwt.client.data.DataSource.getClientOnlyResponse(DataSource.java)
    at com.test.client.RestCacheAllData$1.getClientOnlyResponse(RestCacheAllData.java:51)
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)
    at java.lang.reflect.Method.invoke(Unknown Source)
    at com.google.gwt.dev.shell.MethodAdaptor.invoke(MethodAdaptor.java:103)
    at com.google.gwt.dev.shell.MethodDispatch.invoke(MethodDispatch.java:71)
    at com.google.gwt.dev.shell.OophmSessionHandler.invoke(OophmSessionHandler.java:172)
    at com.google.gwt.dev.shell.BrowserChannelServer.reactToMessagesWhileWaitingForReturn(BrowserChannelServer.java:338)
    at com.google.gwt.dev.shell.BrowserChannelServer.invokeJavascript(BrowserChannelServer.java:219)
    at com.google.gwt.dev.shell.ModuleSpaceOOPHM.doInvoke(ModuleSpaceOOPHM.java:136)
    at com.google.gwt.dev.shell.ModuleSpace.invokeNative(ModuleSpace.java:571)
    at com.google.gwt.dev.shell.ModuleSpace.invokeNativeObject(ModuleSpace.java:279)
    at com.google.gwt.dev.shell.JavaScriptHost.invokeNativeObject(JavaScriptHost.java:91)
    at com.google.gwt.core.client.impl.Impl.apply(Impl.java)
    at com.google.gwt.core.client.impl.Impl.entry0(Impl.java:242)
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)
    at java.lang.reflect.Method.invoke(Unknown Source)
    at com.google.gwt.dev.shell.MethodAdaptor.invoke(MethodAdaptor.java:103)
    at com.google.gwt.dev.shell.MethodDispatch.invoke(MethodDispatch.java:71)
    at com.google.gwt.dev.shell.OophmSessionHandler.invoke(OophmSessionHandler.java:172)
    at com.google.gwt.dev.shell.BrowserChannelServer.reactToMessages(BrowserChannelServer.java:293)
    at com.google.gwt.dev.shell.BrowserChannelServer.processConnection(BrowserChannelServer.java:547)
    at com.google.gwt.dev.shell.BrowserChannelServer.run(BrowserChannelServer.java:364)
    at java.lang.Thread.run(Unknown Source)

You're right though, the earlier crashes I saw listed actual JS calls.
I had tried walking through some of this with the eclipse debugger earlier. Could that be the reason?

I am not an eclipse guru. I know enough to probably get myself in trouble.

Anyway, it looks like the bug's still there.

**Isomorphic** · 1 May 2013, 09:07

If you don't know what running in compiled mode is, please consult the core GWT documentation for Google - it will give you both background and instructions. This is core GWT knowledge.

Also, looks like you haven't consulted the FAQ yet.

**anstaafl** · 1 May 2013, 12:15

OK I have removed the "?gwt.codesvr=127.0.0.1:9997" verified that this works in "production" mode.

I also checked the FAQ "I installed a new version that is supposed to have a fix or new feature, but it's still not working" and followed the instructions in the "full install instructions" link.

About the only thing I was missing was...

<script>
var isomorphicDir = "modulename/sc/";
</script>

Which I added. The full install instructions were for EE so some of the things in it like the copy of log4j.config.xml and server.properties don't apply.

I also deleted the war/<module> and war/WEB-INF/deploy contents before I recompiled and emptied the FireFox cache.

It still fails in hosted/dev mode.

Which begs the question, is it the GWT dev plugin?

If I missed anything could you take pity and throw me a bone here? ;-)

**Isomorphic** · 1 May 2013, 16:36

Did you try use the instructions from the FAQ to verify what version of the library you're loading *in development mode*? Because it sounds very much like you've still somehow got the old version around, but only in Development Mode.

The fact that its still being cached in this mode despite clearing the browser cache would be some kind of bug in GWT. You must try uninstalling the GWT Dev mode browser plugin and re-installing it.

**anstaafl** · 2 May 2013, 08:36

I've reloaded the firefox plugin.

In the test case I last sent you and using the 5/1 libs...

I also deleted everything in gwt-unitcache, war/<modulename> and war/WEB-INF/deploy and recompiled.

In dev mode, firefox crashes when getClientOnlyResponse runs.
IE8 doesn't crash but says the DSResponse getTotalRows from the call says it's null.

For grins, I deleted everything again and added a listgrid to the test case, recompiled it and stuck it on a tomcat server.

On tomcat 7.0.39 With setCacheAllData true...
Both IE8 and FireFox 20.0.1 will display the grid headers and show a "Loading data..." spinner forever.

If I comment out setCacheAllData, recompile and put it on tomcat, I get a listgrid with all the values.

setCacheAllData looks like it's doing the right thing.
the getClientOnlyResponse call is what seems to have issues.
Even after getClientOnlyResponse crashes (or returns null in IE), I can get the cached records with a DS.getCache call.

I've included the updated testcase with the listgrid additions.

Attached Files

**Isomorphic** · 2 May 2013, 08:41

Sorry, there's not a lot we can do for you now. If it's not crashing in compiled mode, you have a staleness issue in Dev mode. We're not sure where it is, so you'll just need to restart/reinstall more things until you find it.

**anstaafl** · 2 May 2013, 08:49

It may not be crashing but it's also not returning any records on a tomcat server when it's not in dev mode.

Anyway. It's just as well.
I think I'll give the beta testing a rest for a while.
Thanks for your help and I do mean that.

**anstaafl** · 4 May 2013, 06:56

One thing I noticed in this stack trace is that the second argument in the getClientOnlyResponse is an array with no closing bracket.

Code:

com.google.gwt.core.client.JavaScriptException: (null) @com.smartgwt.client.data.DataSource::getClientOnlyResponse(Lcom/smartgwt/client/data/DSRequest;[Lcom/smartgwt/client/data/Record;)([Java object: com.smartgwt.client.data.DSRequest@20325350, Java object: [Lcom.smartgwt.client.data.Record;@21215912]): null
    at com.google.gwt.dev.shell.BrowserChannelServer.invokeJavascript(BrowserChannelServer.java:249)
    at com.google.gwt.dev.shell.ModuleSpaceOOPHM.doInvoke(ModuleSpaceOOPHM.java:136)
    at com.google.gwt.dev.shell.ModuleSpace.invokeNative(ModuleSpace.java:571)
...

Shouldn't that be something like...

Code:

@com.smartgwt.client.data.DataSource::getClientOnlyResponse(Lcom/smartgwt/client/data/DSRequest;[Lcom/smartgwt/client/data/Record];)...

At the very least, maybe I helped fix an error message. ;-)

If I can't figure out where this staleness is coming from, I figure I'll download a new version of Eclipse with the latest GWT and your latest libraries and give it one last shot.

Thanks.

**anstaafl** · 6 May 2013, 09:16

I created a new workspace and loaded the testcase into it.

To re-iterate.

Browsers: FireFox 2.0.0.1, IE8
Firefox plugin: 1.20
SmartGWT: SNAPSHOT_v9.0d_2013-05-04 (2013-05-04)
GWT: 2.5.0
Java: version 7 update 21
OS: Windows XP SP3

I downloaded the 5/4/2013 nightly build and ran the test.
The getClientOnlyResponse does return a valid DSResponse object but the data attribute in it is completely empty.

IE8 does not crash but returns "No items to show".
FireFox crashes with the following stack trace...

Code:

12:32:34.294 [ERROR] [RestTest] 12:32:34.281:MOU9:WARN:Log:TypeError: newData is undefined
Stack from error.stack:
    isc.A.fetchRemoteDataReply() @ RestTest/sc/modules/ISC_DataBinding.js:1323
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:240
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:298
    DataSource.fireResponseCallbacks() @ RestTest/sc/modules/ISC_DataBinding.js:640
    DataSource._completeResponseProcessing() @ RestTest/sc/modules/ISC_DataBinding.js:638
    DataSource._handleClientOnlyReply/_6() @ RestTest/sc/modules/ISC_DataBinding.js:501
    DataSource._handleClientOnlyReply() @ RestTest/sc/modules/ISC_DataBinding.js:502
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:240
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:298
    anonymous() @ RestTest/sc/modules/ISC_DataBinding.js:1238
    $wnd.isc.RPCManager.fireReplyCallback() @ :249
    [c]RPCManager.fireReplyCallbacks() @ RestTest/sc/modules/ISC_DataBinding.js:1239
    [c]RPCManager.performOperationReply() @ RestTest/sc/modules/ISC_DataBinding.js:1238
    RPCManager._performTransactionReply() @ RestTest/sc/modules/ISC_DataBinding.js:1231
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:240
    Timer._fireTimeout() @ RestTest/sc/modules/ISC_Core.js:975
    unnamed() @ RestTest/sc/modules/ISC_Core.js:971
    unnamed() @

I have attached the test case (again).

BTW
I think this person http://forums.smartclient.com/showthread.php?t=26358 and myself are seeing the same problem.

Thanks.

Attached Files

**anstaafl** · 6 May 2013, 09:29

Sorry.

I have updated the comments at the top of the file with the correct versions as well as noting that the "Data" attribute in the DSResponse is empty.

Attached Files

RestCacheAllData.java (4.0 KB, 111 views)

**anstaafl** · 6 May 2013, 12:53

Downloaded last night's build so the stack trace would be up to date.

Version SNAPSHOT_v9.0d_2013-05-06 (2013-05-06)

Code:

16:45:56.810 [ERROR] [RestTest] 16:45:56.767:TMR4:WARN:Log:TypeError: newData is undefined
Stack from error.stack:
    isc.A.fetchRemoteDataReply() @ RestTest/sc/modules/ISC_DataBinding.js:1323
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:241
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:299
    DataSource.fireResponseCallbacks() @ RestTest/sc/modules/ISC_DataBinding.js:640
    DataSource._completeResponseProcessing() @ RestTest/sc/modules/ISC_DataBinding.js:638
    DataSource._handleClientOnlyReply/_6() @ RestTest/sc/modules/ISC_DataBinding.js:501
    DataSource._handleClientOnlyReply() @ RestTest/sc/modules/ISC_DataBinding.js:502
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:241
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:299
    anonymous() @ RestTest/sc/modules/ISC_DataBinding.js:1238
    $wnd.isc.RPCManager.fireReplyCallback() @ :249
    [c]RPCManager.fireReplyCallbacks() @ RestTest/sc/modules/ISC_DataBinding.js:1239
    [c]RPCManager.performOperationReply() @ RestTest/sc/modules/ISC_DataBinding.js:1238
    RPCManager._performTransactionReply() @ RestTest/sc/modules/ISC_DataBinding.js:1231
    [c]Class.fireCallback() @ RestTest/sc/modules/ISC_Core.js:241
    Timer._fireTimeout() @ RestTest/sc/modules/ISC_Core.js:976
    unnamed() @ RestTest/sc/modules/ISC_Core.js:972
    unnamed() @

**Isomorphic** · 6 May 2013, 18:41

We see an issue here and we're taking a look

**anstaafl** · 7 May 2013, 03:05

Thank you.

**Isomorphic** · 16 May 2013, 09:06

This should now be fixed, please try the next nightly build.

**anstaafl** · 18 May 2013, 08:40

Excellent!
After verifying the fix with the test case, I loaded it into the 4.0 test version of my full blown app and everything I've done so far has worked as expected.

Thanks again.

BTW
You may want to let this person know as well...

http://forums.smartclient.com/showthread.php?t=26358

Announcement

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment