Red-black tree would spare a machine word per node, because a red-black tree doesn't need depth information.
Hence the reason is either historical or a space/speed trade-off (comparing two depths may be faster than pattern matching).
Regards,
damien guichard
Hi, list,
Just from the curiosity, why balanced binary trees used in Set and Map are AVL-trees, not their alternative, say, red-black trees? Is there a deep reason for it, or just a historical one?