#4 Read-safe concurrency may not work

[Paul Cantrell]

Matthjis Hollemans pointed out to me some weirdnesses with Java's memory model which probably mean that the whole read-safe concurrency idea Eidola uses to minimize synchronization might fail in theory, especially on multiprocessor machines. Here's the explanation:

http://www.javaworld.com/jw-02-2001/jw-0209-double.html

Must look into this....