Andrew John Hughes wrote:
On 3 June 2010 21:12, Alan Bateman <alan.bate...@oracle.com> wrote:
Andrew John Hughes wrote:
:
Thanks Kelly. Your change does make more sense, especially given we
don't yet use the _alpha_ define anyway. I've pushed the revised
version:
http://hg.openjdk.java.net/jdk7/build/corba/rev/2657ee0d2d14
http://hg.openjdk.java.net/jdk7/build/jdk/rev/fb56f86642d6
Thanks,
I realize I'm late to this discussion but I assume the changes to the corba
repo aren't really necessary as it doesn't have any native code any more.
Hmmm, I see this is the case with OpenJDK7 but OpenJDK6 still has
src/share/native/com/sun/corba/se/internal/io/ioser.c. The patch
originated on 6, where presumably the fix was needed.
Yes, I would be open to a backport of Alan's
6939646: Remove obsolete com.sun.corba.se.internal.io package
6955873: CORBA resources bundles and javax.activity missing from b94
to OpenJDK 6.
To be honest, it seems strange that CORBA doesn't just use the JDK
version of Defs-linux.gmk especially when the JDK version is prefixed
with:
+# WARNING: This file is shared with other workspaces.
+# So when it includes other files, it must use JDK_TOPDIR.
+#
Surely these days the only 'other workspaces' to which this applies
are the CORBA and HotSpot ones.
removed some obsolete code there recently but didn't have the cycles to do a
more thorough culling. It's a repo crying out to have its make files
replaced.
Too true, e.g. each compilation warns 'value of JDK_IMPORT_PATH cannot
be empty, check or set ALT_JDK_IMPORT_PATH' when it is perfectly
possible to build without this set.
I was under the understanding that it was maintained as part of
Glassfish so have avoided going near the OpenJDK copy where possible.
What's the situation there?
The copy of corba in the JDK is separately maintained and we don't have
plans to switch to the Glassfish version.
-Joe