In server-side state saving? The 1.1 RI doesn't support simultaneous requests from the same user from two windows; it only works for client- side state saving. The JSF 1.1 algorithm for server-side state saving - just storing the UIViewRoot - necessarily implies that saveState() will not be called, and that's not a bug.
As of JSF 1.2, both server-side and client-side state saving will end up invoking saveState(), because the algorithm for server-side state saving is entirely different in 1.2. I forget how much of this is specified vs. implementation-specific. -- Adam On 3/23/06, Martin Marinschek <[EMAIL PROTECTED]> wrote: > I really wonder how the RI goes about providing a different state tree > for a user if the user has two requests running from the same browser. > > regards, > > Martin > > On 3/23/06, Werner Punz <[EMAIL PROTECTED]> wrote: > > Actually this seems either like a bug in the RI to me, since savestate > > does nothing fancy which should not work in the RI. It basically just > > does what every component does in the save and restore phase. Or a bug > > by the user. > > > > Maybe some of the classes in the original tree are not serializable, > > which could cause a problem, due to the fact that the beans themselves > > are serialized by saveState. > > > > > > > > Mike Kienenberger schrieb: > > > Rick, > > > > > > Sorry. I can't help you. I've never used the JSF RI, nor do I use > > > server-side state saving. > > > > > > You could either implement the special phase listener described in > > > TOMAHAWK-134 or you could try asking on the JSF RI list why > > > saveState/restoreState aren't called for server-side state saving. > > > > > > I'm cc'ing your message to the "MyFaces User mailing list" > > > <[email protected]> as you're far more likely to get help there > > > than emailing me directly. > > > > > > > > > > -- > > http://www.irian.at > > Your JSF powerhouse - > JSF Consulting, Development and > Courses in English and German > > Professional Support for Apache MyFaces >
