topic Re: Error running 5.12 unit tests with Java 7 in Alfresco Archive

Error running 5.12 unit tests with Java 7

brianshowers — Tue, 19 Mar 2013 22:06:29 GMT

Due to the issue with saving TaskEntry objects (originally mentioned here), we are currently using a forked and patched version of Activiti v5.11. I wanted to upgrade to v5.12, so I was going to reapply my patch on top of the 5.12 tag and rebuild. However, I've started getting an error running the

Re: Error running 5.12 unit tests with Java 7

jbarrez — Wed, 20 Mar 2013 10:02:31 GMT

All of our QA runs on JDK 7, and we have only one JDK 6 build to verify: http://jenkins.alfresco.com:8080/jenkins/

Also, when I run it locally with JDK 7 all tests are passing (obviously, or we wouldnt have released) … are you using sun jdk? What platform?

Re: Error running 5.12 unit tests with Java 7

brianshowers — Wed, 20 Mar 2013 15:19:31 GMT

Based on other posts that I had read, that's what I had thought about your test environment. Part of what makes this so confusing.

That said, it certainly appears that in at least some of the 1.7 implementations, Undefined doesn't implement Serializable.

I'm not using anything unusual for the JDK, just Oracle 1.7.0_17 JDK for x64 Windows. I'm running on Windows 7.

Re: Error running 5.12 unit tests with Java 7

brianshowers — Thu, 21 Mar 2013 00:04:02 GMT

Here's another interesting finding. When I execute the test using the 1.7.0_17 i586 (32-bit) JDK, the test passes. So, it appears to me that the Activity build isn't compatible with the 64-bit Java 7 JDK (at least for Windows, I'm not sure about other OS builds).

That said, I have no idea why the implementation of the Undefined class would be different across JDK builds.

Re: Error running 5.12 unit tests with Java 7

jbarrez — Thu, 21 Mar 2013 11:29:37 GMT

Wow, that's a pretty bizar find. I believe our QA is running Red Hat 64 bit. And our systems (OS X) are definitely 64 bit.

I've got no clue why the windows one would be different … let's hope it is a bug and it is fixed in a next release?

That said, I have no idea why the implementation of the Undefined class would be different across JDK builds.

Indeed … that makes no sense….

Altough they are working on getting the new JS engine (Nashorn) into the JDK …

Re: Error running 5.12 unit tests with Java 7

brianshowers — Thu, 21 Mar 2013 22:20:07 GMT

What build number are you using? I just tried with Oracle 1.7.0_17 (64-bit) build for OS X and I ran into the same issue. Undefined doesn't implement Serializable, so the same test fails. There is no 32-bit build for OS X, so I couldn't repeat that part of the scenario.

Re: Error running 5.12 unit tests with Java 7

udoderk — Sun, 24 Mar 2013 21:07:47 GMT

What build number are you using? I just tried with Oracle 1.7.0_17 (64-bit) build for OS X and I ran into the same issue. Undefined doesn't implement Serializable, so the same test fails. There is no 32-bit build for OS X, so I couldn't repeat that part of the scenario.

Hi brianshowers,

Why ever appears sun.org.mozilla.javascript.internal.Undefined ??

P. S Just my two cents: the described problem is not an activiti-problem but Oracle Bug

Re: Error running 5.12 unit tests with Java 7

brianshowers — Mon, 25 Mar 2013 01:54:43 GMT

Hi udoderk,

Thanks for the note! The reason I bring this up here is because it comes into play if you use script blocks in your BPMN diagram such as:


<scriptTask id="script2" scriptFormat="JavaScript" activiti:autoStoreVariables="true">
  <script><![CDATA[
        
    var sum = a + b;           
            
    ]]></script>
</scriptTask>

If you use that functionality, Activiti tries to save the variable "sum" which is of type sun.org.mozilla.javascript.internal.Undefined. The Activiti implementation counts on that class implementing Serializable. If it does not, Activiti cannot save the variable and will throw a runtime exception.

This could very well be a bug in the JDK, but since getting Oracle to fix a bug is not a trivial (or quick) matter, I suspect that the Activiti implementation may need to be changed to work around that bug. On the other hand, it's also possible that in the latest build of the JDK, the Serializable interface was removed on purpose, and it might never come back. Either way, if Activiti is not compatible with particular builds of the JDK, that's something everyone should be aware of.

Re: Error running 5.12 unit tests with Java 7

jbarrez — Mon, 25 Mar 2013 11:22:20 GMT

@Brian: I completely agree with you. I've written down a blog post so people are aware: http://www.jorambarrez.be/blog/2013/03/25/bug-on-jdk-1-7-0_17-when-using-scripttask-in-activiti/

I'll try to find out how to post a JDK bug. However, removing the 'implements Serializable' is something that I can't think that can be done by mistake.

Re: Error running 5.12 unit tests with Java 7

udoderk — Mon, 25 Mar 2013 13:26:01 GMT

Hi,

thank you for this example!

But i assume, that this problem will be occured with last version of JDK6 for 32bit and 64bit machine jdk-6u43
I have downloaded this jdk, then installed it, and then look at jt.jar, located under <jdk1.6.0_43 path>\jre\lib.
This version has no serialization too!

package sun.org.mozilla.javascript.internal;

public class Undefined
{
  public static final Object instance = new Undefined();

  public Object readResolve()
  {
    return instance;
  }
}

i have used jd-gui-0.3.3.windows to display this jar file.
The jdk1.6.0_37 (32bit) has the Serializable Version of Undefined

Re: Error running 5.12 unit tests with Java 7

jbarrez — Mon, 25 Mar 2013 13:56:13 GMT

Hmmm that sure doesn't sound good …

I don't know how we could force the 'undefined' variable to be coerced to a valid java type without depending on an internal sun.org.xxx class..
Maybe it is time to switch to Rhino (https://developer.mozilla.org/en-US/docs/Rhino) as javascript impl for Activiti? The default JDK javascript support has always been wonky and unstable …

Re: Error running 5.12 unit tests with Java 7

brianshowers — Mon, 25 Mar 2013 15:59:31 GMT

@Joram, thanks for the quick action. I have to say that Activiti is, by far, the best open source community that I've worked with. I'm extremely impressed with how quickly you follow up and take action. Most communities tend to de-prioritize any bugs that don't impact their personal code paths.

That said, if there is a way for folks like me to get more involved as a committer, I'm always happy to provide back the fixes and investigations that we find.

Re: Error running 5.12 unit tests with Java 7

jbarrez — Mon, 25 Mar 2013 16:07:22 GMT

@Joram, thanks for the quick action. I have to say that Activiti is, by far, the best open source community that I've worked with. I'm extremely impressed with how quickly you follow up and take action. Most communities tend to de-prioritize any bugs that don't impact their personal code paths.

Thanks!

That said, if there is a way for folks like me to get more involved as a committer, I'm always happy to provide back the fixes and investigations that we find.

We're pretty liberal with committing stuff. The easiest is just creating a pull request on github. That way, it's very easy for us to incorporate any fixes people have made. If you need any directions with how to do that, let me know.
But an investigation like the one you did in this thread is of course also a *very* valuable contribution!

That being said, we still don't have a solution for the jdk 1.7.0_17 problem (and apparently latest jdk 6 now too … ). We're really pondering what would be the best (long-term) solution for this.

Re: Error running 5.12 unit tests with Java 7

jbarrez — Tue, 26 Mar 2013 10:02:26 GMT

Some updates: we discussed it this morning in the team and decided we're gonna make the auto storage of variables not the default anymore. If you want to store something, you need to explicitly call the setVariable method.

For backwards compatibility, we'll keep the 'autoStoreVariables' property so people can set it to true if they really depend on it.

Re: Error running 5.12 unit tests with Java 7

brianshowers — Tue, 26 Mar 2013 13:51:19 GMT

Won't the problem still exist if you try to save a variable of type Undefined? It will be less prevalent since autoSaving will default to off, but I would think that any explicit save operation would result in the same stack trace. Seems like you would also need to give the variables a specific type as well.

That said, I have to admit that we don't actually use script blocks in our diagrams, so I'm certainly no expert in that area.

Re: Error running 5.12 unit tests with Java 7

jbarrez — Tue, 26 Mar 2013 14:17:01 GMT

Won't the problem still exist if you try to save a variable of type Undefined? It will be less prevalent since autoSaving will default to off, but I would think that any explicit save operation would result in the same stack trace. Seems like you would also need to give the variables a specific type as well.

No, when you do execution.setVariable(), the script engine will convert it to a Java type. However, the auto storage does not do that. And we can't do it either cause it are all internal classes of the JDK. So as long as you explicitely call setVariable, all is fine. Also note it is only needed for Javascript (groovy still works ok).

I adapted the default behavior and changed just now on master: https://github.com/Activiti/Activiti/commit/5b1982e719e448b91f2c49d05afbd58f62683666

Re: Error running 5.12 unit tests with Java 7

brianshowers — Tue, 26 Mar 2013 15:05:11 GMT

Got it. Makes total sense.