openjdk team mailing list archive

[Launchpad Bug Tracker <395074@xxxxxxxxxxxxxxxxxx>]

This bug was fixed in the package openjdk-6 - 6b16~pre4-0ubuntu4

---------------
openjdk-6 (6b16~pre4-0ubuntu4) karmic; urgency=low

  [Edward Nevill]
  * Added Bytecode Interpreter Generator.
  * Added ARM templates for above.
  * Removed olde optimised ARM assebler.
  * Added -g0 because of problems with ld linking -g.
  * Changed alignment to 64 now that as bug is fixed.

  [Matthias Klose]
  * Update IcedTea build infrastructure (20090710).
  * Let the -jre package depend on the access-bridge package, not the
    -jre-headless package. LP: #395074.
  * Suggested by Ed Nevill:
    - Pass -timeout:3 when running the jtreg testsuite on zero architectures.
    - Pass -Xmx256M -vmoption:-Xmx256M on armel for the jtreg testsuite run.
  * Tighten build dependency on llvm-dev.

 -- Matthias Klose <doko@xxxxxxxxxx>   Fri, 03 Jul 2009 18:32:50 +0200

** Changed in: openjdk-6 (Ubuntu)
       Status: In Progress => Fix Released

-- 
libaccess-bridge-java-jni recommendation pulls X libraries in a -headless install
https://bugs.launchpad.net/bugs/395074
You received this bug notification because you are a member of OpenJDK,
which is subscribed to openjdk-6 in ubuntu.

Status in “openjdk-6” package in Ubuntu: Fix Released

Bug description:
openjdk-6-jre-headless depends on libaccess-bridge-java, which recommends libaccess-bridge-java-jni, which depends on multiple graphical libraries (libglib2.0-0, libgtk2.0-0...)

This ends up pulling 12Mb of extra libraries on a jre-headless install on a server, and jeopardizes the EucalyptusFitOnCdSpec objectives.

A solution could be to have:
* libaccess-bridge-java SUGGEST libaccess-bridge-java-jni
* openjdk-6-jre RECOMMEND libaccess-bridge-java-jni
That way you would get it installed on -jre and not installed on -jre-headless.

I'm not sure it makes sense from a java-access-bridge perspective, though ?