RE: [picocontainer-dev] NanoWAR and session madness in Tomcat 5

At 10:50 AM 12/30/2005 -0800, you wrote:
 >
 >
 >> Isn't it the ServletContext-container ref that you want to make transient?
 >>
 >> Or then maybe it might be that the session has its reference to it
 >> broken when the sessions are deserialized?
 >
 >Well, I originally thought that.  At least for Tomcat 5, it appears that it
 >calls "SessionBound" for every session that is deserialized from state.  So
 >at least, the container proper would have its construction capabilities
 >restored. (And in fact, it appears that the new container then overwrites
 >the old one).
 >
 >I don't know how it would work for clustering, but I agree with Mauro that
 >if we can at least get single-machine working with Tomcat, then I'm ok with
 >postponing investigation of clustering until post 1.0.
 >

We already know that it works with Tomcat in at least one case; using the 
groovy pico syntax rather than the builder syntax (maybe not with the 
context container, because I don't store a component in the context 
container, but with the request container, which we also know fails 
serialization when using the builder syntax, so I presume it's the same 
problem being discussed).  And that's with no changes whatsoever to 
ServletContextListener.  If modifying ServletContextListener makes it all 
work with all syntaxes, that would be great.  However, the fact that it 
works with the pico syntax with no changes means that the change would 
simply be masking a more fundamental issue that should probably be solved 
first.

Jake

 >                                                       -Mike (R)