Working with computers DOES damages memory.
I remember when I first came upon a computer... mmm... well, actually I dont
remember...
:-)
Shahar.
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]]On Behalf Of Lennart Petersson
Sent: Wednesday, January 03, 2001 3:05 PM
To: jBoss
Subject: Re: [jBoss-User] TX question
Yep! When i left the room for lunch i hit me... there was another thread
about this for not so long ago. Bad memory. Sorry!
/Lennart
Rickard ?berg skrev:
>
> Hi!
>
> Shahar Solomianik wrote:
> > Perhaps you should invoke the overloaded methods using
> > getEJBContext().getEJBObject().getNextKey.... and not directly.
> > As I understand, that way the container is aware of the method
invokation
> > (cause its done through the remote interface, actually) and the
transaction
> > is handled correctly.
>
> Yes, that's it.
>
> /Rickard
>
> >
> > Shahar.
> >
> > -----Original Message-----
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED]]On Behalf Of Lennart Petersson
> > Sent: Wednesday, January 03, 2001 1:00 PM
> > To: jBoss user
> > Subject: [jBoss-User] TX question
> >
> > Having a CMP bean UniqueKey.
> >
> > Among all methods there are following:
> >
> > public String getNextKey()
> > //This methods returns getNextKey(xxx)
> >
> > public String getNextKey(String upperLimit)
> > //This methods return getNextKey(xxx,yyy)
> >
> > public String getNextKey(String upperLimit, String incrementValue)
> > //This method calculates a new unique key
> > //Calls setLastKey(newKey)
> > //Returns newKey
> >
> > public void setLastKey(String newKey)
> > //Updates the persisted field with newKey
> >
> > The client is in this case calling on the getNextKey() method with no
> > argument, so the whole chain is used.
> >
> > I've used ejbdoclet 0.71 to generate all but the bean class and using
> > Oracle 8.1.5 as database
> >
> > The bean has TX_SUPPORTS and my first try was to only have
> > TX_REQUIRES_NEW on setLastKey().
> >
> > Result: Nothing updated in DB but the bean gives me a new key every
> > time.
> >
> > Next try was to have TX_REQUIRES_NEW on getNextKey().
> > Result: DB updated and beans gives me a new key every time.
> >
> > Note that it had no affect if a switched to TX_REQUIERES_NEW on one of
> > the other getNextKey(...) methods, it was the one that was primarly
> > called that was important.
> >
> > Is this really correct or a bug? In my logic it should be ok to have
> > TX_REQUIRES_NEW only on the last method in the chain, setLastKey()?
> >
> > /Lennart
> >
> > --
> > mailto:[EMAIL PROTECTED]
> > http://www.benefit.se/english
> >
> > --
> > --------------------------------------------------------------
> > To subscribe: [EMAIL PROTECTED]
> > To unsubscribe: [EMAIL PROTECTED]
> > List Help?: [EMAIL PROTECTED]
> >
> > --
> > --------------------------------------------------------------
> > To subscribe: [EMAIL PROTECTED]
> > To unsubscribe: [EMAIL PROTECTED]
> > List Help?: [EMAIL PROTECTED]
>
> --
> Rickard ?berg
>
> Email: [EMAIL PROTECTED]
>
> --
> --------------------------------------------------------------
> To subscribe: [EMAIL PROTECTED]
> To unsubscribe: [EMAIL PROTECTED]
> List Help?: [EMAIL PROTECTED]
--
mailto:[EMAIL PROTECTED]
http://www.benefit.se/english
--
--------------------------------------------------------------
To subscribe: [EMAIL PROTECTED]
To unsubscribe: [EMAIL PROTECTED]
List Help?: [EMAIL PROTECTED]
--
--------------------------------------------------------------
To subscribe: [EMAIL PROTECTED]
To unsubscribe: [EMAIL PROTECTED]
List Help?: [EMAIL PROTECTED]
