Bugzilla – Full Text Bug Listing
| Summary: | Mapped Feature available versions node shows all IU versions from all mapped repos | ||
|---|---|---|---|
| Product: | [Technology] CBI | Reporter: | Miles Daffin <miles.daffin> |
| Component: | CBI p2 Repository Aggregator | Assignee: | Project Inbox <b3.aggregator-inbox> |
| Status: | CLOSED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | thomas |
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Windows XP | ||
| Whiteboard: | |||
|
Description
Miles Daffin
There are five different locations where we can search for versions available: 1. in the parent repository. 2. from other enabled repositories in the current validation set 3. from enabled repositories in other validation sets 4. from disabled repositories. If we show more than #1, then we need a good way to visualize the different sources. Here's one suggestion: - The versions from #1 and #2 will appear like they do today. - The versions from #3 will be grayed out. - The versions from #4 will not be visible at all. - All visible versions will be annotated with a match/don't match indication. The normal versus grayed out appearance signifies what is included or not in the validation. The reason for leaving out #4 is that a disabled repository is normally not loaded (it is only loaded if it has been enabled sometime since you opened the editor). Including versions under such circumstances would be confusing. #4 should definitely not appear. As you say, the repo is not loaded when disabled. (If I disabled a repo I would expect to see the versions contributed by it to disappear from all available version lists too.) #3 would it make sense for versions in a sub-set to appear in the super set (the set that extends and therefore I suppose includes the other set)? How easy would it be to differentiate between #1 and #2? What threw me initially was seeing versions I did not think were in a particular repo. I had to check the repo browser to make sure. (And when I disabled other repos I thought might be contributing versions the list did not update.) I would be nice (not essential) to be able to tell if a version came from another repo. Fixed in rev. 1534. OK. Fix indicates where a version comes from using a comment in parentheses. (Additionally, text for versions coming from other mapped repos is greyed out.) # If version comes from the same mapped repo: no comment necessary (#1). # If version comes from a different mapped repo in the same contribution comment is "from other Repository in the Contribution". # If version comes from a different contribution in the same validation set comment is "from other Contribution in the Validation Set". # If version comes from a different validation set comment is "from other Validation Set". # If version comes from a disabled mapped repo - we don't see it. This is getting a bit convoluted I think. There are 2 possible refinements: 1. Change the text in the comments to make it simpler (see below) 2. Replace the comments with the model path to the version: Validation Set label/Contribution label/Mapped Repo label. Here are some suggested minor changes to the label text: # If version comes from a different mapped repo in the same contribution comment is "from another Repository in this Contribution". # If version comes from a different contribution in the same validation set comment is "from another Contribution in this Validation Set". # If version comes from a different validation set comment is "from another Validation Set". As discussed, we do not expect to see versions from mapped repos that have been disabled. This is the case if the mapped repo is disabled when we load the model into the editor. But if we disable the mapped repo its contributed versions do not disappear from the available versions lists. We either have to update the version range in the mapped feature or close and reopen the model. Likewise, if we delete a mapped repo its contributed versions do not disappear from the available versions lists. (This looks tricky to get right.) Noticed that comment text is not always grey. When is it supposed to be grey? This suggests that the version in questions comes from a mapped repo that is disabled. Is this what is meant? (In reply to comment #6) > Noticed that comment text is not always grey. When is it supposed to be grey? Entries stemming from other ValidationSets are grey since they are not actually available during the validation. They are just there for reference. I changed the texts. OK. Changes verified. One question. When should the list of available versions update? At the moment I can disable/enable and add/remove stuff but the list does not update until I open/close the model. They should be updated immediately. Please open a separate bugzilla if that doesn't happen. Closing this one. [Bookkeeping change only. Moving bugs to the new "home" of aggregator, CBI. Made no changes to assignee's for closed bugs, even though some were old inbox.] |