archie at whistle.com
Mon Feb 22 10:20:37 PST 1999
Godmar Back writes:
> I need to read up on this, but I vaguely remember somebody saying that
> synchronized is no longer part of the spec. That is, if a function
> is not thread-safe, it must protect itself somehow, but that's not part of the
> interface. Should that turn out to be true (somebody check this), then
> yes, let's omit synchronized --- but only where it's immediately obvious that
> it's not needed.
My contribution on this question is the observation that at least
"javadoc" has taken this stance. In 1.2, the "synchronized" attribute
is no longer part of the javadoc documentation for a method.
So I think this would be OK. I agree however that "only where it's
immediately obvious that it's not needed".
Archie Cobbs * Whistle Communications, Inc. * http://www.whistle.com
More information about the kaffe