On 2/7/2011 7:17 PM, Niclas Hedhman wrote:
On Tue, Feb 8, 2011 at 9:29 AM, Patricia Shanahan<p...@acm.org> wrote:
Perhaps it would help understanding if you gave your reasons for not going
to Java 6 at this time?
Maybe I missed it; What is the driver for going Java 6?
It has taken many other projects a long time when reaching the "need"
to go Java 6 and dropping Java 5 support.
I would rather like to be able to drop Java 5 support because we could
then use some features in java.util and its sub-packages. However, I see
that as a nice-to-have, nothing I can't program around, and I'm
concerned about the impact on users.
The Java 1.4 drop is a different issue. I found that I could break a key
outrigger class, FastList, by sufficiently intense stress testing. I
have so far failed to identify the bug, but it can be fixed relatively
easily by substituting a new, much simpler, cleaner FastList based on
ConcurrentLinkedQueue and AtomicLong. If we can't Java drop 1.4 I have
to go back to debugging aggressively concurrent code with the old memory
model :-(
Patricia
