#opendaylight-meeting: tws

Meeting started by phrobb at 17:07:18 UTC (full logs).

Meeting summary

    1. edwarnicke walks quickly through Fast/Phased (phrobb, 17:09:36)
    2. rovarga brings up the issue of bug tracking across branches (skitt, 17:14:50)
    3. JIRA fixes this, colindixon also describes processes to handle it in bugzilla (skitt, 17:16:56)
    4. about semantic versioning, edwarnicke brings up the issue of skew and how to collapse it (skitt, 17:18:15)
    5. rovarga explains how to use version ranges to handle that (skitt, 17:18:48)
    6. e.g. a dependency on [1.0.0,2) automatically pulls in the latest release in that range (skitt, 17:19:31)
    7. this works for features too (skitt, 17:19:38)
    8. for this to work people need to adhere to semver (skitt, 17:19:53)
    9. first step is to see if we can cascade a Maven release throughout the project instead of relying on autorelease's single reactor (skitt, 17:27:08)
    10. rovarga thinks this won't work at first but it will give us a list of things to fix (skitt, 17:27:25)
    11. edwarnicke points out that for this to work everywhere we need to get whatever change we want to make to propagate throughout the project (skitt, 17:28:31)
    12. possible for plan for Boron: switch odlparent and yangtools to semantic versioning (skitt, 17:33:48)
    13. and use version ranges for their downstreams (skitt, 17:33:58)
    14. animal sniffer can point out missing semantic version bumps (skitt, 17:34:43)
    15. the question ends up being the human part: how do we involve the projects to avoid version skew (skitt, 17:51:13)
    16. how do we avoid projects blocking upstream version bumps (skitt, 17:51:23)
    17. how do we handle large version bumps when blockages do occur (skitt, 17:51:38)
    18. colindixon points out that there are two orthogonal things under discussion: semver and phased (skitt, 17:54:05)
    19. few projects will have the discipline for semver in the short term (skitt, 17:54:16)
    20. but we can reap the benefits broadly without involving everyone (or can we?) (skitt, 17:54:35)

Meeting ended at 18:06:50 UTC (full logs).

Action items

  1. (none)

People present (lines said)

  1. skitt (20)
  2. odl_meetbot (4)
  3. phrobb (3)
  4. CaseyODL (1)
  5. edwarnicke (0)

Generated by MeetBot 0.1.4.