Jon Zeppieri wrote:
>
> Artur,
>
> The 1.2 javadocs on Hashtable (but not HashMap, strangely) suggest that
> a conforming implementation *must* be bucket-based:
>
> "Note that the hash table is *open*: in the case [of] a 'hash
> collision', a single bucket stores
> multiple entries, which must be searched sequentially."
As long as it is not in Java Spec, we can safely ignore such comments.
Remeber that it should be not possible to determine implementation
externally - and it is true in case of Hashtable.
Using such class would have one additional advantage - it would be very
different from sun implementation :)
Artur
