Discussion:
[werkflow-user] blog example not firing publish/reject process
Federico Spinazzi
2003-12-04 16:27:30 UTC
Permalink
Hi,
I'm trying werkflow.
I've just read the article 'The application of Petri Nets to Workflow
Management' and starting to play with werkflow (beta-6) for an
application using JavaGroups as the messaging transport.

As I've some problems with the blog demo (I'm running it from IntelliJ)
and seen on the mailing list archives both the old messages and the cvs
commit.

The problem is that the reject and publish processes are not invoked.

From stepping in the debugger it seems to me like the message is asked
for acceptance but never consumed (??).

Anyone has the same problem?

Thank you very much
Federico
Jason van Zyl
2003-12-04 16:47:07 UTC
Permalink
Post by Federico Spinazzi
Hi,
I'm trying werkflow.
I've just read the article 'The application of Petri Nets to Workflow
Management' and starting to play with werkflow (beta-6) for an
application using JavaGroups as the messaging transport.
As I've some problems with the blog demo (I'm running it from IntelliJ)
and seen on the mailing list archives both the old messages and the cvs
commit.
The problem is that the reject and publish processes are not invoked.
From stepping in the debugger it seems to me like the message is asked
for acceptance but never consumed (??).
Anyone has the same problem?
I would look at the ActivityManagerTest:

http://cvs.codehaus.org/viewcvs.cgi/werkflow/werkflow-core/src/java/test/org/codehaus/werkflow/initiator/message/?root=codehaus

If you're embedding werkflow in an application this is most likely the
idiom of usage that will be most helpful. This test also works so is a
good starting place.
Post by Federico Spinazzi
Thank you very much
Federico
_______________________________________________
werkflow-user mailing list
http://lists.codehaus.org/mailman/listinfo/werkflow-user
--
jvz.

Jason van Zyl
***@zenplex.com
http://tambora.zenplex.org

In short, man creates for himself a new religion of a rational
and technical order to justify his work and to be justified in it.

-- Jacques Ellul, The Technological Society
Mark H. Wilkinson
2003-12-04 17:18:32 UTC
Permalink
Post by Jason van Zyl
Post by Federico Spinazzi
As I've some problems with the blog demo (I'm running it from IntelliJ)
and seen on the mailing list archives both the old messages and the cvs
commit.
The problem is that the reject and publish processes are not invoked.
From stepping in the debugger it seems to me like the message is asked
for acceptance but never consumed (??).
Anyone has the same problem?
Yes, I see the same problem. I thought the example had been fixed, but
it seems not to be working as I'd expect it to.
Post by Jason van Zyl
http://cvs.codehaus.org/viewcvs.cgi/werkflow/werkflow-core/src/java/test/org/codehaus/werkflow/initiator/message/?root=codehaus
If you're embedding werkflow in an application this is most likely the
idiom of usage that will be most helpful. This test also works so is a
good starting place.
Hmm; that test is making sure that a message will initiate a process,
but it doesn't check that a subsequent message will correlate to the
existing process and wake it up again. I think we need to extend the
test and see if the there's something here that needs fixing.

-Mark.
Jason van Zyl
2003-12-04 17:47:14 UTC
Permalink
Post by Mark H. Wilkinson
Post by Jason van Zyl
Post by Federico Spinazzi
As I've some problems with the blog demo (I'm running it from IntelliJ)
and seen on the mailing list archives both the old messages and the cvs
commit.
The problem is that the reject and publish processes are not invoked.
From stepping in the debugger it seems to me like the message is asked
for acceptance but never consumed (??).
Anyone has the same problem?
Yes, I see the same problem. I thought the example had been fixed, but
it seems not to be working as I'd expect it to.
Post by Jason van Zyl
http://cvs.codehaus.org/viewcvs.cgi/werkflow/werkflow-core/src/java/test/org/codehaus/werkflow/initiator/message/?root=codehaus
If you're embedding werkflow in an application this is most likely the
idiom of usage that will be most helpful. This test also works so is a
good starting place.
Hmm; that test is making sure that a message will initiate a process,
but it doesn't check that a subsequent message will correlate to the
existing process and wake it up again. I think we need to extend the
test and see if the there's something here that needs fixing.
Definitely, I just think this test represents the most common use case
as far as usage goes. Even if people don't think so yet :-)
Post by Mark H. Wilkinson
-Mark.
_______________________________________________
werkflow-user mailing list
http://lists.codehaus.org/mailman/listinfo/werkflow-user
--
jvz.

Jason van Zyl
***@zenplex.com
http://tambora.zenplex.org

In short, man creates for himself a new religion of a rational
and technical order to justify his work and to be justified in it.

-- Jacques Ellul, The Technological Society
Federico Spinazzi
2003-12-10 12:46:06 UTC
Permalink
Post by Jason van Zyl
Post by Mark H. Wilkinson
Post by Jason van Zyl
http://cvs.codehaus.org/viewcvs.cgi/werkflow/werkflow-core/src/java/test/org/codehaus/werkflow/initiator/message/?root=codehaus
If you're embedding werkflow in an application this is most likely the
idiom of usage that will be most helpful. This test also works so is a
good starting place.
Hmm; that test is making sure that a message will initiate a process,
but it doesn't check that a subsequent message will correlate to the
existing process and wake it up again. I think we need to extend the
test and see if the there's something here that needs fixing.
Definitely, I just think this test represents the most common use case
as far as usage goes. Even if people don't think so yet :-)
Ehmm,
I've seen recent commits on cvs so maybe you are working on this (now
some test fails during the build).
In either case thanks for your attention and time.
Federico Spinazzi

Loading...