Sébastien Lorion <[EMAIL PROTECTED]> wrote: > Considering that operations on the tree are centralized in the Tree class,
By making this assertion, IMO you've pre-selected a particular notion of interacting with a tree. Usually, when I work with a tree, my node *is* my tree (i.e. it's the root of the tree). What's more, I often don't store parent pointers and instead rely on e.g. visiting algorithms to store the parent stack during a recursive traversal, if such parents are required for certain operations. And taking a leaf (ahem) from the pages of functional programming, when parent pointers are eliminated, subtrees may be shared between multiple trees, since my nodes are often read-only after construction (and thus aliasing issues aren't important). There's more down this branch too: persistent datastructures often rely on the logarithmic performance of operations on trees to permit keeping e.g. undo-state around quite cheap. Can you say more about your particular tree? -- Barry -- http://barrkel.blogspot.com/ =================================== This list is hosted by DevelopMentor® http://www.develop.com View archives and manage your subscription(s) at http://discuss.develop.com
