Well, the cost is more of shifting the current pojo at index (i+1) to (i) with detach and then attach when I do a remove (o) where o resides at index (i). So I am saying we don't have to shift since Set is neither ordered or indexed. BTW, just looking at it again, I don't think add(o) will prevent duplication as key can be different. What do you think? Yeah, the whole suite is functionalPojoCacheTests. It'd certainly be nice to run through it once a while. :-) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3984405#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...