OH and BTW,
I didn't document that but I firmly believe that the cache "acquisition"
should be somewhat different than "cache synchronization" right now it is
mixed in there. It is the sync that is the issue.
In other words, go on as if nothing happened Simone, i will probably factor
the sync behaviour in an interceptor right above it...
regards
marc
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Bordet, Simone
> Sent: Tuesday, August 08, 2000 6:54 AM
> To: jBoss Development Mailing List (E-mail)
> Subject: [jBoss-Dev] Entity Instance Cache
>
>
> Hello,
>
> I'm rewriting the instance cache for entity beans, and in
> NoPassivationEntityInstanceCache I see a remark done few days ago
> that will
> avoid throwing a RemoteException when a reentrant call is made and isn't
> allowed.
> Marc, can you explain me in more detail what happened and why you remarked
> it ?
>
> Code follows:
>
> // MF FIXME
> // This is wrong but doing it right requires time
> // The problem is that the entity EJB calls from an instance
> // come back on the instance and in the presence of a
> // transaction it throws the exception....
> // Since I suspect most people will use the EJBObject method calls before
> they set the reentrant right :)
> // I would rather bypass the default for now.
> //throw new RemoteException("Reentrant call not allowed");
>
> I will try to fix this doing it right, but I don't follow you in the
> comments.
>
> Thanks
>
> Simon
>
>
