Chris Toshok writes:
> > CVS checkout as of today
> > I am not up to date on your code base.
> you aren't? please update before sending bug reports...
I said: CVS checkout as of today. I just haven't read and
understood every single cvs-commit message from the last
month's backlog.
> I thought I actually responded to you about your static test relying on
> behavior that's undefined in the JLS. I can't remember the exact text
> of my response, and don't have a copy :(
There was a list debate about initialization in general.
I voted for "initialize on load", which is not in
accordance with the specs.
I do not remember any discussion of the "static" test
specifically. However, the simple test I wrote down
is just a stupid way to code. The inheritance between
StaticSuper and StaticSub serves no purpose, and
without, the problem does not occur. If that's what
you pointed out in a mail I missed, I see your point.
> by all means, use what works.
There is seemingly nothing that just *works* on Linux,
once JNI and the Invocation come into play. The project
I was investigating for has just abandoned Java. Well,
that means that my life got a lot easier, as I do not
have to stick to the Invocation API any longer,
which caused the majority of problems.
b.
