| Summary: | [planner] Have the explanation return all the missing requirements | ||
|---|---|---|---|
| Product: | [Eclipse Project] Equinox | Reporter: | Pascal Rapicault <pascal> |
| Component: | p2 | Assignee: | P2 Inbox <equinox.p2-inbox> |
| Status: | ASSIGNED --- | QA Contact: | |
| Severity: | enhancement | ||
| Priority: | P3 | CC: | dj.houghton, irbull, katya.stoycheva, leberre, s.yousouf |
| Version: | 3.7 | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Mac OS X - Carbon (unsup.) | ||
| Whiteboard: | |||
|
Description
Pascal Rapicault
I guess we could simply collect all the missing requirements and return them if any. Note that in that case we would not have a real explanation, i.e. we would not get the path from the request to the missing requirements, only the missing reqs. Else we could compute a real explanation. When you say we could collect all the missing requirements. How would that differ from what we have in Slicer? If we do something in that space, and if at all possible, I think we should strive to compute a real explanation. I was thinking about the #missingRequirement() method in the projector. The problem with real explanations, is that we need to return a set of explanations instead of a single explanation currently, which will probably break the API. |