In the Wagn 2.0 models, a simple Deck is a namespace, or a set of bindings. Another way of saying this is that it is a set of key value pairs, much like a hash in many languages. In the wagn model, the keys are Card Names and the values are Card Content. The name is not properly part of the card content, but even if <card>+name is made to mean "the name of this card", you can't just change it, you have to change the binding to the namespace. Cards may be bound into different namespaces, so <card>+name could represent a sort of "home namespace binding" even when the card appears through other namespace mappings. <card>+namespace would be needed too in foreign namespaces.
We use set language to call up the mathematical models of sets. Any projection of a namespace that also meets the properties of a namespace can be used as a namespace. Pointer and search cards, or "all the plus cards recursively from a card" are also card sets and can be projected into different namespaces by features of the Deck of Cards model space.
I want to distinguish a strong form of namespaces and a more general one that can support name related features inside of a deck. The strong form implies that all the parts of all names within the namespace, but this will not be true of this weaker sense. In this weaker sense, any cardname is a namespace for all cards below it, and recursively any namespaces it contains, and we can test if a card is in a namespace, including or excluding any subspaces stacked below it.