v0.1.0

This release is backward compatible with the previous one.

Changed

• Added ThreadMapX type that contains a Mutex for each value, allowing the methods Self::fold, Self::fold_values, and Self::probe to run more efficiently when there are concurrent calls to the per-thread methods (Self::with and Self::with_mut) by using fine-grained per-thread locking instead of acquiring an object-level write lock. On the other hand, the per-thread methods may run a bit slower as they require the acquision of the per-thread lock.