Is "cannot" a matter of being contrary to policy, or does including such references break the repository to the point where installations will fail?
I would agree that code-wise this is an issue with java-package, however, the work required to bring java-package into line is several orders of magnitude greater. The sun-java6-jre package (IIRC) involves downloading the binary from Sun after the dummy package is installed, so _inclusion_ of the packages in repositories is not an option. The two methods currently use conflicting numbering schemes - java-package generates 1.6.x version numbers, while the Ubuntu dummies are numbered 6.0.x. Because of the differences, an "upgrade" from one to the other will cause problems, so a new package name will need to be devised - I am guessing that this in itself is a scary process. Notwithstanding the package naming issues, I will file this against java-package too. I would attempt a fix myself, but last time I read those scripts my head exploded. FWIW, I believe I may have successfully run 2.5 (or was it 2.4?) under gij/gcj, but this was long enough ago that I can't reliably recall. Will test this at a later date. Any alternative solutions in the interim (short of evil --force options) that would allow the upgrade? -- Missing dependency on java2-runtime et al. https://bugs.launchpad.net/bugs/174198 You received this bug notification because you are a member of Ubuntu Bugs, which is the bug contact for Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
