Delivery precondition - no unused imports firing against new work and not checked in files (changeset)
![]() One of my users noted an odd behavior in 602 ifix10 using SCRUM. They have a deliver pre-op trigger "no unused imports on delivery" which I believe is an out-of-the-box setup.
The problem noted is as follows:
- In their workspace they modified a few files and checked them in, creating a changeset and linking it to a work item. This now shows up in the users Outbound folder in the RTC client.
- The user then continued development of more work in the same workspace but has not checked them in yet.
- User then tried to deliver the Outbound change but is stalled by the pre-opt trigger on "No unused imports". the user stated that the checked in code did not have any.
It appears that the pre-opt deliver trigger is scanning the users current workspace (containing work in progress) and not the checked in changeset.
Has anyone noted this behavior? Looks like a potential bug.
Norm Dignard
|
Accepted answer
![]()
Geoffrey Clemm (30.1k●3●30●35)
| answered Jul 23 '17, 9:32 a.m.
FORUM ADMINISTRATOR / FORUM MODERATOR / JAZZ DEVELOPER As Ralph indicates, the client-side preconditions that refer to an Eclipse semantic status (such as no errors or no unused imports) are about the Eclipse status, and not the status of just the change-sets being delivered. If you want to ignore some your pending changes, you could check those changes into a change set (if they aren't already checked-in), then suspend those change sets, execute your deliver, and then resume those suspended change sets. Norman Dignard selected this answer as the correct answer
|
One other answer
![]()
Ralph Schoon (63.3k●3●36●46)
| answered Jun 26 '17, 3:46 p.m.
FORUM ADMINISTRATOR / FORUM MODERATOR / JAZZ DEVELOPER edited Jun 27 '17, 6:19 a.m. This is the intended behavior. The operation is a client side behavior and unused imports only makes sense in the context of a local project that is local on disk. The unused imports and organize imports is an Eclipse capability and does not know or care about change sets. The operation only prevents you from delivering if there are unused imports. You have to fix it and can potentially only deliver additional changes that fix the unused imports. Comments The problem here is that he is not trying to deliver work in progress (not checked in code). He's trying to deliver code that has already been checked in and that does not have unused imports, etc.. The error/warning is false. ![]() FORUM ADMINISTRATOR / FORUM MODERATOR / JAZZ DEVELOPER
This is the behavior since RTC 1.x and before. This s the first time in 9 years that I see a complaint.
The user can basically
Note that the advisor to prevent delivery in case of compiler errors in the local files has the same behavior and exists for the same reason:
|
Your answer
Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.