This is the mail archive of the mailing list for the Java project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: need to focus on java performance?

Andi Vajda writes:
 > On Sun, 21 May 2006, Tom Tromey wrote:
 > > FWIW this isn't that interesting of a goal for me.  I think
 > > compatibility is very important.  For one thing it is the way to
 > > leverage all the existing free java code out there.  The idea is, we
 > > have a free system from bottom to top, except for this one proprietary
 > > slice in the middle.  And our goal, as I see it, is to make this slice
 > > free as well.
 > Does this mean that, if Sun were to fully open source Java so that
 > it were really free, the gcj project would be dropped ?

Well, that depends.  There surely would be less interest in gcj, but
it's free software: as long as people want to maintain it it will
never die.

 > While a really free java would be great, losing gcj as a
 > consequence would be very sad. In addition to providing a free java
 > environment, gcj makes java eminently more usable by non-java
 > programs. It breaks Java's insularity by making it available from
 > C++ and other languages via CNI, embeddable, and a lot smaller when
 > I'm able to statically link a shared library with libgcj which I
 > can do with gcj 3.4.x).

Right.  I agree with all that.

The fact that gcj works with all the standard system tools re
debugging, profiling, etc, is a tremendous advantage.  In the
proprietary world you have Java profiling and instrumentation tools
separate from OS-level tools.  This is stupid.

But right now it is extremely hard for gcj to compete effectively,
because we have so few people working on it.  Red Hat has been
extremely generous with its support for this project, but that alone
may not be enough.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]