This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: Java interface to Xlib : JCNIX
- From: Tom Tromey <tromey at redhat dot com>
- To: "R. A. Rivas Diaz" <rivasdiaz at yahoo dot com>
- Cc: java at gcc dot gnu dot org
- Date: 05 Nov 2002 12:49:13 -0700
- Subject: Re: Java interface to Xlib : JCNIX
- References: <20021105194711.98128.qmail@web21403.mail.yahoo.com>
- Reply-to: tromey at redhat dot com
>>>>> "Rivas" == R A Rivas Diaz <rivasdiaz@yahoo.com> writes:
Rivas> Is anyone working on a AWT implementation? I want to help in
Rivas> this but I don't know where to start.
Someone on the Classpath list recently volunteered to work on AWT.
But, the more the merrier...
There are several things that can be done.
The Java side of AWT is known to be buggy. Parts are missing (e.g.,
GridBagLayout, plus lots of java2d, etc). Parts don't work properly
(we don't use the tree lock anywhere). Lots to do here. Look here:
ftp://alpha.gnu.org/pub/gnu/classpath/nightly/tests/classpath-jdk13-compare.txt
This page compares Classpath against the JDK and can help you figure
out what to do.
The peers have problems too. When I've done AWT work I've used the
Gtk+ JNI peers in Classpath. I think this remains the best approach
for the time being. Completing and debugging the peers is very
worthwhile. It would also make sense to rewrite the peers to use CNI
(though I see that as lower priority).
Setting things up to work on AWT using libgcj isn't very hard. I can
repost the instructions if you want.
If you're going to work on this, it is best to get started on the
paperwork as soon as possible (if you've already done this and I
forgot, apologies). Also it makes sense to coordinate with the
Classpath hackers, since AWT is a fully-merged package.
Tom