[JBoss Tools] New message: "Re: JBossTools-3.0.3.GA On Eclipse Ganymede"
by Salim OFLAZ
User development,
A new message was posted in the thread "JBossTools-3.0.3.GA On Eclipse Ganymede":
http://community.jboss.org/message/527810#527810
Author : Salim OFLAZ
Profile : http://community.jboss.org/people/salimoflaz
Message:
--------------------------------------------------------------
Hi Denis,
Thank you for your reply. I already tried seam generate entities, but it didn't work also. I am using HSQL database in single user mode. It didn't do anything. Some debug messages to console is printed from hibernate code generation facility such as cannot find any columns and cannot find primary key. I guess it's due to naming of database objects such as schema name and table names. Do you have any experience with code generation facility with HSQL database? I am thinking of migrating my db to mySQL. Do you have any opininon? Or should I post a new thread for this question? Thanks,
Best Regards,
Salim OFLAZ
--------------------------------------------------------------
To reply to this message visit the message page: http://community.jboss.org/message/527810#527810
16 years, 2 months
[Javassist] New message: "Re: javassist.bytecode.BadBytecode: unset variable. Please Help Me!"
by jaikiran pai
User development,
A new message was posted in the thread "javassist.bytecode.BadBytecode: unset variable. Please Help Me!":
http://community.jboss.org/message/527799#527799
Author : jaikiran pai
Profile : http://community.jboss.org/people/jaikiran
Message:
--------------------------------------------------------------
> megalodon wrote:
>
> Some more information regarding this error:
>
> 1. In prefix method, I am creating a new instance and assigning it to localMIName variable. Later I am just printing the newly created object (with over ridden toString() method). So there are 2 statements here. One to assign new instance to localMIName and another to print localMIName value. *IF, these 2 statements are submitted to CtMethod.insertBefore() function as a single block, I get no exception!* Mind you, If I add a statement later on to print localMIName in a separate statement (another CtMethod.insertBefore()) call, I still get the exception! What happens to the assigned value? Does it go out of scope?? Why does this happen only in JSP classes?
>
>
>
I haven't looked much into the details and can't help much as to what's wrong with your code. But looking at some of your posts here, you seem to be having a hard time trying to debug these javassist runtime issues. Sometime back, there was thread in this forum where a user pointed to a open source library which he had developed as an wrapper to Javassist http://www.fuin.org/srcgen4javassist/index.html . In one of his examples http://www.fuin.org/srcgen4javassist/examples.html he shows how it can be used to generate this debug information from javassist. I haven't personally given it a try, but maybe you might find it useful. You could use that library to wrap to javassist code and print out debug information - thay way you can see the exact code that has been generated.
--------------------------------------------------------------
To reply to this message visit the message page: http://community.jboss.org/message/527799#527799
16 years, 2 months
[jBPM] New message: "How to handle rollback due to runtime errors"
by Maciej Swiderski
User development,
A new message was posted in the thread "How to handle rollback due to runtime errors":
http://community.jboss.org/message/527792#527792
Author : Maciej Swiderski
Profile : http://community.jboss.org/people/swiderski.maciej
Message:
--------------------------------------------------------------
Hi all,
in one of recent posts I have been interesting discussion going on about exception handling in jBPM. As it was described there user code exception should be handled within that code but what about case where user code was executed fine but something else went wrong?
Please consider following scenario:
- there are 4 nodes (regardless of type) that are executed as one transaction.
- some of the nodes modify external systems via different protocols
- no exceptions are thrown from nodes
- runtime exception is thrown due to some db issues
Transactionwill be rolled back but changes made to external system are not rolledback since they were not part of the transaction (for example webservice call).
Have you put any thoughts into that? I was wondering if some sort of compensation mechanism could be introduced to the execution where nodes that take part in transaction can inform runtime engine what to invoke in case of roll backs. This could give kind of undo operation for already finished nodes and should not cause any problems when process will go thru them again.
What's your opinion about it? Does it sound reasonable?
Cheers,
Maciej
--------------------------------------------------------------
To reply to this message visit the message page: http://community.jboss.org/message/527792#527792
16 years, 2 months