oki thanks! So my fix is ok I guess. Still I'm not sure if I like it to be forced to have each and every @SessionScoped and @ConversationScoped bean to be serializable!
Usually it's a good idea for clusters etc, but otoh there are some good arguments against it in some cases. But thats a different story and should be discussed on the EG list. LieGrue, strub --- Gurkan Erdogdu <[email protected]> schrieb am Fr, 13.3.2009: > Von: Gurkan Erdogdu <[email protected]> > Betreff: Re: problems with @Serializable > An: [email protected] > Datum: Freitag, 13. März 2009, 20:10 > Proxy objects are always Serializable > in the specification > > > > > ________________________________ > From: Mark Struberg <[email protected]> > To: [email protected] > Sent: Friday, March 13, 2009 9:02:20 PM > Subject: problems with @Serializable > > > I'm facing a really naughty problem while I try to add the > > implements Serializable > > to all @SessionScoped test beans: > > First I thought that Javassist has a bug with reading some > _valid_ classes because I found a few sources which insist > on using cglib instead of javassist in this case > [1][2][3][4] > > but after digging deeper I found the following line in the > JavassistProxyFactory#63: > > > interfaceList.add(Serializable.class); > > Why do we need this? Are proxy objects always > serializable? > > I will check in a workaround by simply checking if > Serializable already exists, but would be happy if you can > review this Gurkan! > > txs and LieGrue, > stru > > > [1] http://opensource.atlassian.com/projects/hibernate/browse/HHH-1938 > > [2] http://www.jboss.org/index.html?module=bb&op=viewtopic&t=149575 > [3] http://opensource.atlassian.com/projects/hibernate/browse/HHH-3719 > [4] https://jira.jboss.org/jira/browse/WBRI-122 > > >
