At 09:39 AM 21/11/2003 -0800, Boehm, Hans wrote:
>The use of finalizers may be motivated by native code, but
>often it won't be.
This is true, but I thought problem at issue wasn't finalization per se,
but premature finalization.
Clearly, putting ones finalization in the wrong class can mess things up,
but no amount of synchronization is going to fix inherently broken code.
Maybe I've completely misunderstood the issue, but is there an example of
premature finalization that does not use a native resource, and which
cannot be dealt with by moving finalization processing into the correct class?
Sylvia.
-------------------------------
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel
This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:54 EDT