hi ludovic, please also test the latest snapshot (i've pushed a small improvement).
fyi: your issue is that myfaces-core is doing full state-saving in this case. -> so it's quite essential (for your application) that you find the issue/constellation. @scope issue: that sounds more like an issue in combination with myfaces-core 2.2.1 (or the cdi-implementation your are using). regards, gerhard http://www.irian.at Your JSF/JavaEE powerhouse - JavaEE Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2014-03-13 8:55 GMT+01:00 [email protected] <[email protected]>: > On 13/03/2014 08:45, [email protected] wrote: > >> On 12/03/2014 09:16, Gerhard Petracek wrote: >> >>> hi ludovic, >>> >>> i've pushed the fallback for full state-saving. >>> however, please provide the demo which illustrates the issue (it might >>> be an unexpected issue in myfaces-core). >>> >> Hi. >> >> I spent some time trying to prepare a simple demo, but could not >> duplicate the bug in a simple case. Sorry. >> >> I suspect a issue related to injection of a SessionScoped bean in the >> ViewAccessScoped bean, but lack a simple example. >> >> I (or my colleague Laurent cottereau) might come back soon with a maybe >> related issue, maybe on MyFaces list. We noticed a strange behaviour with >> @ViewAccessScoped beans. >> >> It seems that, under some conditions, they are not properly >> instantiated/destroyed but rather recycled. Moreover, it seems that we do >> not access the "right" bean in @PostConstruct or @PreDestroy. In fact, it >> seems that we access a different instance of what should be the same bean >> when in a call of an EL expression and in @PostConstruct/@PreDestroy. >> > Oh and this bug occurs with : > - MyFaces 2.2.1 > > and > - CODI 1.0.5 > or > - DS 0.6-SNAPSHOT (201401312 snapshot) > > Working on a simple test... > > Best regards, > > > Ludovic > | > | AVANT D'IMPRIMER, PENSEZ A L'ENVIRONNEMENT. > | > >
