In looking at Joseph's suggestions, I realized that there was a
problem with the existing and proposed description of sleep. The
problem is that the existing and proposed text interacts badly with
NPT and currentTimeMillis(). The way it was written, if a thread
tried to sleep for ten seconds, and during that that period NTP
caused the clock to be rolled back one hour, the thread would sleep
for one hour. We would prefer for that not to happen, although we
can't guarantee it if we don't have access to a clock that is immune
to NTP changes.
I changed the spec to more closely follow the Javadoc:
>Causes the currently executing thread to sleep (temporarily cease execution)
>for the specified duration, subject to the precision and accuracy
>of system timers and schedulers.
>The thread does not lose ownership of
>any monitors, and resumption of execution will depend on scheduling
>and the availability of processors on which to execute the thread.
>
>A sleep for a period of zero time and yield operations need not have
>obversable
>effects.
Bill
-------------------------------
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel
This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:57 EDT