Community
Participate
Working Groups
I'm trying to remove p2 feature from console: osgi> provlgp org.eclipse.cvs.feature.group 1.3.100.v20101029-7B77FO-8sF7B7AHCR5BQ75 org.eclipse.cvs.source.feature.group 1.3.100.v20101029-7B77FO-8sF7B7AHCR5BQ75 org.eclipse.egit.feature.group 0.8.4 org.eclipse.equinox.p2.user.ui.feature.group 2.1.0.v20101103-897OFx-FdHjO2NlKwKOHf8E org.eclipse.equinox.p2.user.ui.source.feature.group 2.1.0.v20101103-897OFx-FdHjO2NlKwKOHf8E org.eclipse.help.feature.group 1.3.0.v20101118-7i7tFE9FFt6ZopgSqKRnz0tH org.eclipse.help.source.feature.group 1.3.0.v20101118-7i7tFE9FFt6ZopgSqKRnz0tH org.eclipse.jdt.feature.group 3.7.0.v20100824-0800-7z8eFcgFMTfEQG1byPJlI_d9H155 org.eclipse.jdt.source.feature.group 3.7.0.v20100824-0800-7z8eFcgFMTfEQG1byPJlI_d9H155 org.eclipse.jgit.feature.group 0.8.4 org.eclipse.mat.feature.feature.group 1.0.100.201012150941 org.eclipse.pde.feature.group 3.7.0.v20100803-7b7oFXNFEx2XmfiPhi5933 org.eclipse.pde.source.feature.group 3.7.0.v20100803-7b7oFXNFEx2XmfiPhi5933 org.eclipse.platform.feature.group 3.7.0.v20100923-9fF7MHDqFsAkplGz0n61z-yU57WHoz0JdMfaI org.eclipse.platform.source.feature.group 3.7.0.v20100923-9fF7MHDqFsAkplGz0n61z-yU57WHoz0JdMfaI org.eclipse.rcp.feature.group 3.7.0.v20101115-9FB-FqhFr3P05j0S-HRVgSR org.eclipse.rcp.source.feature.group 3.7.0.v20101115-9FB-FqhFr3P05j0S-HRVgSR org.eclipse.sdk.feature.group 3.7.0.v20101118-7Q7xA7DPa2xdN86nbxlooGNr6aWgeOpuz-xq-WrZxV_3Q org.eclipse.sdk.ide 3.7.0.I20101208-1300 org.eclipse.team.svn.feature.group 0.7.9.I20100512-1900 osgi> provremove org.eclipse.egit.feature.group 0.8.4 Remove complete for org.eclipse.egit.feature.group 0.8.4 After provremove reported successful completion I tried confapply, restart of dropins, restart of eclipse to apply the changes, but the feature is still on the system. The profile still has the feature I'm trying to remove. I also tried with other features like "org.eclipse.cvs.feature.group 1.3.100.v20110520-0800-7B78FHk8sF7BB7SAB4GVG5", but with the same result - the feature cannot be removed.
I tested this and it seems that the message being displayed on completion of the uninstallation is incorrect. For example if I try to uninstall the org.eclipse.equinox.p2.user.ui.feature.group it says it succeeds but it can't because it is included in the SDK. However if I install something that I had installed on top of my SDK it gets uninstalled correctly. Would you mind looking into this issue? The code is in org.eclipse.equinox.p2.console.
It is strange, however, that "provremove org.eclipse.sdk.ide" worked perfectly - errors started to pop up and after I stopped the IDE, eclipse.exe was gone. On the other hand, "provremove org.eclipse.equinox.p2.user.ui.feature.group" doesn't seem to uninstall the feature - it is still listed with provlgp. In a debug session we saw that both the calculation and execution of the provisioning plan finished with status OK. Is this correct ? How are the planner and the engine are supposed to know not to touch features from the SDK ? As a further note, I noticed that the creation of the provisioning plan is immediately followed by its execution. Since the creation produces a status result, shouldn't it be checked first before proceeding with the execution ?
> On the other hand, "provremove org.eclipse.equinox.p2.user.ui.feature.group" > doesn't seem to uninstall the feature - it is still listed with provlgp. In a > debug session we saw that both the calculation and execution of the > provisioning plan finished with status OK. Is this correct ? How are the > planner and the engine are supposed to know not to touch features from the SDK? The behaviour you are observing matches the semantics we gave to uninstall. Uninstall is attempting to remove what is being identified, but this may result in a no-op if the thing being uninstalled is still depended upon. For example if I install GMF (which depends on GEF and EMF), then install EMF (which is a no-op since EMF is already installed), then go and uninstall EMF, I don't want to remove GMF because of this. From a user point of view, the fact that GMF depends on EMF is an implementation detail. If needs be the planner could be changed to return an INFO status indicating that nothing will be done (though technically speaking there may be things that will be done like removing the root flag from the profile). Another semantics that could have been given to uninstall is one where we force the IU to be uninstalled and fail if it is depended upon. However I don't think this is a very suitable model for a user facing functionality. > As a further note, I noticed that the creation of the provisioning plan is > immediately followed by its execution. Since the creation produces a status > result, shouldn't it be checked first before proceeding with the execution ? The status is checked in the first line of PlanExecutionHelper#executePlan
> Another semantics that could have been given to uninstall is one where we > force the IU to be uninstalled and fail if it is depended upon. However I don't > think this is a very suitable model for a user facing functionality. Still this could be added (it is basically requesting a complete negation of the IU you are trying to uninstall)
This bug hasn't had any activity in quite some time. Maybe the problem got resolved, was a duplicate of something else, or became less pressing for some reason - or maybe it's still relevant but just hasn't been looked at yet. If you have further information on the current state of the bug, please add it. The information can be, for example, that the problem still occurs, that you still want the feature, that more information is needed, or that the bug is (for whatever reason) no longer relevant. -- The automated Eclipse Genie.
