> The trouble is that the lifecycle of a Wicket WebSession object is not the > same as the HTTPSession. When a session is temporary only, you never get a > valueUnbound() callback. I'll have to move the resource manager > initialization into the valueBound() method; inconvenient but not a problem > per se.
Ah yeah, but temporary sessions are volatile anyway. You could implement a custom request cycle that in onEndRequest checks whether the session is temporary, and if so, let it call the cleanup method. > My second objection is that this works, as you said, with the default > implmentation of ISessionStore. I would introduce a dependency on the exact > implementation of ISessionStore used, not any published API. That somehow > fills me with sombre premonitions ;-) > I still think it would be nice if org.apache.wicket.Session had a proper > lifecycle (à la WebApplication.onDestroy()). Should I file a feature request > or start a conversation on the dev list? Sure, you can start a discussion on the dev list, and filing a JIRA issue makes that it will be a more persistent discussion topic. Having it here is fine too though. I'm not sure whether depending on the Session object being put in the HttpSession is bad for you though. If you are not using that, it probably means you're working with a custom session store in which case you could just make sure your cleanup callback gets called by that, right? Eelco --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
