Hello, I have been silently following the recent discussion on the new CM and have a question about what it means for a component to be thread safe.
In my mind I feel that thread safe is being mixed up with "stateless". In other words, a "stateless" component implies that it is thread safe, but the relation is not if and only if! I can imagine a component that is perfectly thread safe but not stateless. With a stateless component, it make sense that cm.lookup() will always return the same component, but if the component is statefull, then it should always return a new component (or pooled.) So with the new CM, do we have to assume that everything returned from lookup() is stateless and use a factory interface for stateful components? If we make the statefullness of a component a property of the metadata, then the CM can at least be smart enough to return a shared or new instance of the component. I am thinking along the lines of a Session EJB. One thing I like about the EJB design is that an EJBObject has a remove() method and the stateful vs stateless behavior is a property of the metadata. This design makes it easy to develop a generic container/factory for Session beans. Now, in the Avalon world, what if remove() was added to the Component interface and the CM looked for some attribute in the component's config to determine its statefulness? Regards, --mike -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
