It is a bit daunting and possibly impossible as there is enough gray
area in the spec
as to implement a language that is not capable of passing the Java
validation suite of tests and
that is the only thing I can think of that could be used to verify
the user has implemented
the spec. (hmmm... a bit circular).
Kaffe is a clean room implementation to the spec but they have expanded
the
language in some interesting ways. I seem to recall they used
the MS alternative
to JNI but maybe they put in JNI as well. From a recent read of Kaffe
I got the
impression that they stopped at 1.1.1 and did not know if they were
headed for 1.2
or 1.3.
Of course once we enter the standards process we can do anything from
start with
a clean sheet of paper up to (more or less) Java.
I presented a paper on using JNI with Linux at the Atlanta Linux Showcase
and the feedback
ran from one guy telling me this was exactly what he needed to someone
else wanting to
rewrite the entire Java class library with JNI and opensource.
A replacement library that
relied on opensource would be a good place to start and I even had
a publisher interested in
a book along those lines.
What next? Well I guess I'll start making some calls and get back
to those who are
interested and start a splinter group.
Nathan Meyers wrote:
Tony Dean wrote:
> 1) Sun owns the Java trademark. They have published the VM spec and
> the language spec. They permit rogue ports from the specs.There is already an excellent "rogue port" in the Kaffe project,
although "cleanroom implementation" is a better term. Interestingly,
even the spec carries some scary language... check out paragraph 6 of
the copyright from the JVM spec, about the license to "practice this
specification":http://java.sun.com/docs/books/vmspec/2nd-edition/html/Copyright.doc.html#997057
Nathan
-- Tony Dean Linux: The choice of a GNU Generation!
