Community
Participate
Working Groups
Steps: 1. Enable xmlrpc on a repository that does not support XML-RPC Exception Stack Trace: java.lang.IllegalArgumentException at org.eclipse.core.runtime.ListenerList.add(ListenerList.java:103) at org.eclipse.core.internal.jobs.InternalJob.addJobChangeListener(InternalJob.java:158) at org.eclipse.core.runtime.jobs.Job.addJobChangeListener(Job.java:162) at org.eclipse.mylyn.internal.bugzilla.core.BugzillaClientManager.repositoryChanged(BugzillaClientManager.java:111) at org.eclipse.mylyn.internal.tasks.core.TaskRepositoryManager$5.run(TaskRepositoryManager.java:467) at org.eclipse.core.runtime.SafeRunner.run(SafeRunner.java:42) at org.eclipse.mylyn.internal.tasks.core.TaskRepositoryManager.notifyRepositorySettingsChanged(TaskRepositoryManager.java:459) at org.eclipse.mylyn.internal.tasks.core.TaskRepositoryManager$1.propertyChange(TaskRepositoryManager.java:77) at org.eclipse.mylyn.tasks.core.TaskRepository.notifyChangeListeners(TaskRepository.java:830) at org.eclipse.mylyn.tasks.core.TaskRepository.setProperty(TaskRepository.java:823) at org.eclipse.mylyn.internal.bugzilla.ui.tasklist.BugzillaRepositorySettingsPage.applyToInternal(BugzillaRepositorySettingsPage.java:710) at org.eclipse.mylyn.internal.bugzilla.ui.tasklist.BugzillaRepositorySettingsPage.applyTo(BugzillaRepositorySettingsPage.java:559) ...
I also got an error popup "Can not get the Default Milestons using XMLRPC".
I hope to have an fix soon. The problem is that when finish is pressed we did not validate the settings!
Steffen, shoud i use the New gerrit instance for this fix? Hope that i can do this later Today.
That's up to you. If it's a bigger a change or you would like some feedback it's a good idea to push it to Gerrit but if you are confident that the change does not require a review you can also push it directly to the Eclipse.org.
Steffen, I use gerrit (http://review.mylyn.org/27) to make sure that we have no failing tests. Can you please review so we are sure that my approach is OK.
Created attachment 203083 [details] mylyn/context/zip
We have a failing test! org.eclipse.mylyn.trac.tests.core.TracTaskDataHandlerXmlRpcTest.testAttachmentChangesLastModifiedDate I think that this is not related to my changes!
No worries, I have already run the clean up script which has fixed the failing test.
I have commented on the review: http://review.mylyn.org/#change,27 . We'll need to improve the validation a bit before we can push this.
(In reply to comment #9) > I have commented on the review: http://review.mylyn.org/#change,27 . We'll need > to improve the validation a bit before we can push this. I put my comment!
I see two new reviews: http://review.mylyn.org/#change,45 http://review.mylyn.org/#change,46 Can you push an update to the original one at http://review.mylyn.org/#change,27 instead?
(In reply to comment #11) > I see two new reviews: > http://review.mylyn.org/#change,45 > http://review.mylyn.org/#change,46 > > Can you push an update to the original one at > http://review.mylyn.org/#change,27 instead? Sorry, I only want to push http://review.mylyn.org/#change,47 for bug#354023 (the XMLRPC part)
(In reply to comment #10) > (In reply to comment #9) > > I have commented on the review: http://review.mylyn.org/#change,27 . We'll > need > > to improve the validation a bit before we can push this. > > I put my comment! Did you publish the comments? Initially comments are only drafts but once you select "Review" and then "Publish Comments" the comments become visible for others as well.
(In reply to comment #13) > (In reply to comment #10) > > (In reply to comment #9) > > > I have commented on the review: http://review.mylyn.org/#change,27 . We'll > > need > > > to improve the validation a bit before we can push this. > > > > I put my comment! > > Did you publish the comments? Initially comments are only drafts but once you > select "Review" and then "Publish Comments" the comments become visible for > others as well. Sorry I did not know this! Now you can see my comments!
Thanks. We should either check if XML-RPC is available transparently when making an XML-RPC call or store that as part of the configuration. It also makes sense to check that as part of validation but there shouldn't be another job that does verification.
Steffen, I start from scratch and create patch set 2 for review 27. Can you review this?
Created attachment 206184 [details] mylyn/context/zip
I'll try to take a look later this week but the change looks quite large so it might take some time.
(In reply to comment #18) > I'll try to take a look later this week but the change looks quite large so it > might take some time. I found some failing Tests so wait until i have a fix for this
*** Bug 362314 has been marked as a duplicate of this bug. ***
Steffen, i create patch set 4
Created attachment 206549 [details] mylyn/context/zip
Thanks. Some of the problems from the last review are still present in the latest patch set. The entire logic for validation XML-RPC connectivity should be encapsulated in BugzillaClient.validate() which should return an object that describes the discovered properties for the Bugzilla installation (version, XML-RPC available, etc.).
Concerning an extra control for reporting validation results on the settings page: I would prefer if we came up with a generic solution in the framework (bug 219680).
(In reply to comment #24) > Concerning an extra control for reporting validation results on the settings > page: I would prefer if we came up with a generic solution in the framework > (bug 219680). Steffen, I have comment on the review!
Frank, I reverted commit 0d08a0917251974ba720a2bd2828f19f252e3b7f for now. The spirit of the change seems fine but the API proposal needs more work and discussion before we can proceed with that. I have opened this bug to track the enhancement that I think your change implemented. We have often discussed that as a desirable feature enhancement on conference calls: 370331: automatically validate when pressing finish on repository setting page https://bugs.eclipse.org/bugs/show_bug.cgi?id=370331
This was fixes with bug#370331
