Grouper Product Backlog

Current Tasks

1. [GRP-12] Provide more efficient means of retrieving all child groups and/or stems (short)

2. [GRP-6] Fix subject caching (medium)

3. [GRP-10] Modify membership indices (short)

4. [GRP-11] StemFinder#internal_isChild generates spurious error logging (short)

5. [GRP-22] Minimize member queries (medium)

6. [GRP-19] Refactor privilege resolution (medium)

7. [GRP-24] Reorder privilege resolvers to restore prior behavior (short)

8. [GRP-7] Simplify UI for browsing (medium)

9. [GRP-21] Add Group attribute caching (short)

10. [GRP-17] Release Grouper API 1.2.1

Grouper Product Backlog

Most of these tasks are currently not sufficiently specified to provide an estimate on how long they will take to complete.

1. [GRP-18] Update Ldappc to use better child group query filter (short)

2. [GRP-14] Poor integration with Ldappc: Null DTO error message in Ldappc run causes run to fail (medium)# [GRP-15] Alter query workflow to restrict by scope before searching (medium)

3. Integrated QuickStart package (? days)
4. Changes/history (? days)
5. Subject API v1.0 (? days)
6. Multi-valued group attributes (5 days)
   • blair@uchicago.edu: This will require a schema change. By adding three new columns (klass, attribute_type and is_system) to the grouper_attributes table the API could support a) multi-valued group attributes, b) simplify the grouper_stems table, c) simplify the grouper_groups table, d) possibly simplify the grouper_composites table, e) treat group type assignments as normal attributes and f) allow fields to be assigned to more than one group type. I think. Not all of these changes would need to be performed immediately but modifying the schema appropriately would allow the rest of these changes to be made on an incremental basis.
     1. Do we need to restrict whether attributes are single- or multi-valued on a per attribute basis? If so, the complexity grows.
     2. Another option would be to a) make the schema change so as to allow all of secondary benefits but b) not allow normal group attributes to be multi-valued at this time.
7. Add change history support to ldappc (? days)
8. Subject API v1.1 (? days)
9. People Picker (? days)
10. Standardized Java and web services interfaces (? days)
11. Signet-Grouper combined UI (? days)
12. Aging & reactivation (? days)
13. Hooks (? days)
14. Namespace transition (? days)
15. Audit (? days)
16. Integrate with collaboration platform (? days)
17. Multiple groupers (? days)
18. Grouper should work with Hibernate 3 (4 days)
19. DRY "create()", "delete()" and "update()" in DAO (2 days)
20. Relocate GrouperTest out of the public API (1 day)
21. Fields can be associated with more than one group type (5 days)
22. Add SAX-based XML import tool (15 days)
23. Do not hand generate XML during export (15 days)
24. Web services interface for querying and management (? days)
25. Assume immediate='true' in XML import (3 days)
26. Allow full Subject tuple to be specified in XML export-and-import tools (2 days)
27. Allow subject id to be used instead of subject identifier in XML export-and-import tools (2 days)
28. Add special logic to SubjectFinder.findById(id) and SubjectFinder.findByIdentifier(idfr) for GrouperAll and GrouperSystem (1 day)
29. Don't write privileges to database when granting privileges to "GrouperSystem" in default privilege adapters (4 days)
30. Add DAO dependency injection (10 days)
    Source: blair@uchicago.edu
31. Review error logging so that faults, but not contingencies, are logged (2 days)
    Source: blair@uchicago.edu
32. Database ERD (1 day)
    Source: James Cramton @ Brown
33. Dynamic Groups (30 days)
34. Some way of honoring FERPA in the API and UI (30 days)
    Source: Shilen Patel @ Duke
35. Make subject id of current grouper session available in the Subject API (30 days)
36. Improved Integration Between I2MI Components (30 days)
    Source: blair@uchicago.edu
    • blair@uchicago.edu: This could be as simple (to begin with) as adding an "install" target to the Grouper API, Grouper UI , the Subject API and Ldappc. Perhaps even gsh? A common directory structure would also help.

      Questions or comments?  Contact us.