|
Bugzilla – Full Text Bug Listing |
| Summary: | zypper lists duplicated solvables | ||
|---|---|---|---|
| Product: | [openSUSE] openSUSE 11.0 | Reporter: | Thomas Göttlicher <tgoettlicher> |
| Component: | libzypp | Assignee: | Jan Kupec <jkupec> |
| Status: | RESOLVED DUPLICATE | QA Contact: | E-mail List <qa-bugs> |
| Severity: | Enhancement | ||
| Priority: | P5 - None | CC: | ma |
| Version: | Beta 3plus | ||
| Target Milestone: | --- | ||
| Hardware: | Other | ||
| OS: | Other | ||
| Whiteboard: | |||
| Found By: | --- | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
|
Description
Thomas Göttlicher
2008-05-26 14:50:47 UTC
Not only xmlout, but also normal out. This is intended. Zypper is not GUI, where you can click to open the Versions tab, where you will see all the available versions, together with the duplicates form different (or duplicated) repositories. This is not a simple thing to solve and there are several related issues: * duplicated repositories (obviously the way to go here is to avoid them - different bug/feature) * duplicated solvables from _different_ repositories. What to do with them? Filter them out? Probably not. When/how to display them (or feed to xml). * 'lu' cannot show selectables instead of solvables, as there can be different versions of a patch wich can both be satisfied or both needed (how does YaST handle this?) IMO the major pain here are the duplicated repositories. Other stuff is OK. So i'd close this as duplicate of bug #377568, but first what are others' opinions? (In reply to comment #1 from Ján Kupec) > IMO the major pain here are the duplicated repositories. Other stuff is OK. So > i'd close this as duplicate of bug #377568, but first what are others' > opinions? We had to change the organization of our pool. Unfortunately this is a very deep change. We currently load all packages (items) from all repos and treat each item individually. We had to unify same items and link each item to multiple repos providing it. 1) we must define when 2 items are the the same. (e.g same sha1 sum of the binary rpm, and something reasonable for patterns/products). Same items must producde the same solver result (so must have at least same name, edition, arch, vendor, dependencies)... 2) the pool then has to unify same packages. One of them would be the provider for any metada, and instead of a repository attribute, each item would contain a list of repositories. This is something we could do if there are many repositories with overlapping (but not identical) content. But IMO too much effort if we just deal with duplicate repositories (same content). *** This bug has been marked as a duplicate of bug 400895 *** |