OpenJDK 6 Skeleton Plan for OSX Universal Binary
neugens.limasoftware at gmail.com
Fri Oct 29 08:23:32 PDT 2010
First of all, sorry if this mail breaks your in-thread mail client
formatting, but I just subscribed to the list.
I was redirected here from a comment on my blog:
I was asked few times during the last couple of years how feasible
would be to have a native OSX port of the AWT peers, and I noticed
that someone proposed the JKoala project, so let me cite what is
written on their web site:
How long should porting AWT / Swing take?
I estimate the porting task will take between one and two years.
Depending on other developers who might join JKoala project meanwhile,
this delay could be greatly reduced.
In all cases, expect regular news and test versions to let you follow
JKoala development. It's the way open source communities generally
I completely agree that a full port, full in the sense of complete
compatibility etc... including the native look and feel will probably
take more than one year and a little less than two, but we do have
some short cuts that will enable us a faster and usable
implementation, giving us enough time to actually refine all the
As someone may know, I am the maintainer of the Caciocavallo project,
we won the OpenJDK challenge (actually, we classified 3th place) by
proposing a full refactoring of the AWT peers. Once finished, we
developed a full AWT stack on top of Swing. The port is finished and
feature complete, although there are bugs of course.
We use it internally and we have all sort of peers for Cacio, you can
find some example on those blogs:
Plus some post I can't find about Cacio on Microware OS 9 :)  and
Windows CE that we showcased at JavaOne this year.
An OpenJDK based VM with working Swing and AWT (and Java2D) has been
delivered, or is in the process of being delivered, on a multitude of
operating systems, with good performances, despite the huge class
library and the fact that most of those platforms are mobile.
I don't have any real experience of OSX coding (but once again, I
didn't have any previous experience on those funny platform as well),
but I'm sure that an initial port (provided that the JDK itself
already compiles and works, perhaps with the X11 based peers), will
only require few weeks of work (assuming real work).
I would like to jump in and help, although I don't have a Mac yet.
Cacio needs to be fixed to be used on the latest code base, because we
use an older drop of the OpenJDK code, but I expect this to be quite
trivial (and definitely something I can do on my Linux machine).
I don't know if there is any way I can help further and I have some
constraints (if you read my blog, you'll know about those shortly),
but if you are interested in sponsoring such a porting effort somehow,
I'll be very help to help.
pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
Fingerprint: BA39 9666 94EC 8B73 27FA FC7C 4086 63E3 80F2 40CF
Proud GNU Classpath developer: http://www.classpath.org/
Read About us at: http://planet.classpath.org
Please, support open standards:
More information about the bsd-port-dev