On Wed, 4 Sep 2002 18:25, Leo Sutic wrote: > I was more thinking that: > > That is, "conn-manager" implies: > > + A certain interface (org.apache.avalon.ConnManager) > > + A set of attribute keys (for example, it may declare "is-secure"), > but it does not imply the actual values of those keys. That is, > the "conn-manager" implies a list of attribute names, but not > the values of those attributes.
So equivelent to Cocoons (and myrmidons) short-name? Essentially just a short name that maps to the Service (or Role in Cocoon/myrmidon terminology). So that "conn-manager" is a name for the type. > I'm all for the third case, where the role string has container scope, > or > the first case where it has consumer scope. The second case I find > harder > to accept. I just see a load of naming issues. Okay - I get you. The first case is the one I have been proposing. However that differs from what I interpreted you to mean just above. The key (whether it be "abcdefg" or "harry-potter" is local to and defined by the consumer. It is the containers responsibility to map a component into the component local namespace (ie using the "abcdefg" or "harry-potter" keys). I guess what I am saying is that we can have a shortname "conn-manager" that refers to the service type. I believe Stephen did this at one stage for blocks using their "name" field. However this is not necessarily related to the name of the key that consumers use to get a reference to the component. -- Cheers, Peter Donald *-----------------------------------------------------* | Never argue with an idiot, they'll drag you down to | | their level, and beat you with experience | *-----------------------------------------------------* -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
