[concurrency-interest] Re: AtomicInteger and AtomicLong should
Sun, 04 Jan 2004 21:08:05 -0600
Larry Riedel wrote:
>>It seems really odd to not have [AtomicInteger and
>>AtomicLong] implement Number.
> Let me count the ways Java is odd or just plain lame. I would
> like to imagine the new stuff tries not to be encumbered by
> antediluvian cruft like java.lang.Number, which is a Class,
> not an Interface, and thus kind of Useless, thanks to the Java
> precept that multiple inheritance of implementation is Evil,
> leaving us with objects and "containers" duct taped together
> with byzantine naming conventions, "dynamic Proxy"s, etc.
You know, I actually never noticed that Number was actually an abstract
Okay, so let me ask a different question. Why can't this JSR fix this
discrepency and change Number to an interface and provide a new abstract
base class if that is still appropriate?
If all we do in JSRs is create new code and never fix the old cruft,
then how will things ever get better?