)]}'
{"openstack_releases/cmds/validate.py":[{"author":{"_account_id":12898,"name":"Tony Breeds","email":"tony@bakeyournoodle.com","username":"tonyb"},"change_message_id":"6aed9679fd3d3014548f406ca2efee66c7fa7432","unresolved":false,"context_lines":[{"line_number":108,"context_line":"            # An empty file results in None, so convert to dict to"},{"line_number":109,"context_line":"            # make using the value easier."},{"line_number":110,"context_line":"            previous_deliverable \u003d {}"},{"line_number":111,"context_line":"    for branch in previous_deliverable.get(\u0027branches\u0027, []):"},{"line_number":112,"context_line":"        if branch[\u0027name\u0027] \u003d\u003d expected_branch:"},{"line_number":113,"context_line":"            # Everything is OK"},{"line_number":114,"context_line":"            return"}],"source_content_type":"text/x-python","patch_set":2,"id":"3a461143_f4b473d5","line":111,"updated":"2017-02-02 02:24:39.000000000","message":"I like this but given the current data in the release repo, I think this will block a heap of releases.   For example a release of nova from the newton branch.  IIUC given there are no branches defined in deliverables/mitaka/nova.yaml (even though the branch clearly exists) we\u0027ll never go through this loop always hitting the mk_error() call.\n\nRather than add branches statements everywhere should we also check if the branch exists in git before calling mk_error()?\n\nAs an aside if we *did* add the branches statements everywhere in the near future does that imply we can remove diff-start key (can\u0027t we intuit it from $previous_series/branches/location ?)","commit_id":"3d71ba11ee17af237f73fc97d3d3517812861b5a"},{"author":{"_account_id":2472,"name":"Doug Hellmann","email":"dhellmann@redhat.com","username":"doug-hellmann"},"change_message_id":"2525a6c0eefaff7405f3b034f102e8ca74493494","unresolved":false,"context_lines":[{"line_number":108,"context_line":"            # An empty file results in None, so convert to dict to"},{"line_number":109,"context_line":"            # make using the value easier."},{"line_number":110,"context_line":"            previous_deliverable \u003d {}"},{"line_number":111,"context_line":"    for branch in previous_deliverable.get(\u0027branches\u0027, []):"},{"line_number":112,"context_line":"        if branch[\u0027name\u0027] \u003d\u003d expected_branch:"},{"line_number":113,"context_line":"            # Everything is OK"},{"line_number":114,"context_line":"            return"}],"source_content_type":"text/x-python","patch_set":2,"id":"3a461143_bd863211","line":111,"in_reply_to":"3a461143_1ced968e","updated":"2017-02-02 18:07:59.000000000","message":"Rather than rewrite this, I moved the call below into the block that is only invoked for the \"current\" series, and then I rebased the patch on top of the one that opens pike by setting the default to pike. That should mean the rule only applies from pike forward, which is the situation I was originally concerned about.","commit_id":"3d71ba11ee17af237f73fc97d3d3517812861b5a"},{"author":{"_account_id":2472,"name":"Doug Hellmann","email":"dhellmann@redhat.com","username":"doug-hellmann"},"change_message_id":"5e654cf1a550b9286734d1b6057e8a92ae08ab72","unresolved":false,"context_lines":[{"line_number":108,"context_line":"            # An empty file results in None, so convert to dict to"},{"line_number":109,"context_line":"            # make using the value easier."},{"line_number":110,"context_line":"            previous_deliverable \u003d {}"},{"line_number":111,"context_line":"    for branch in previous_deliverable.get(\u0027branches\u0027, []):"},{"line_number":112,"context_line":"        if branch[\u0027name\u0027] \u003d\u003d expected_branch:"},{"line_number":113,"context_line":"            # Everything is OK"},{"line_number":114,"context_line":"            return"}],"source_content_type":"text/x-python","patch_set":2,"id":"3a461143_1ced968e","line":111,"in_reply_to":"3a461143_f4b473d5","updated":"2017-02-02 13:50:08.000000000","message":"Good call. I\u0027ll see if I can make it look at the branches in git. I\u0027ll have to move the call below to after a point when the repo exists, I think.\n\nNice idea about dropping diff-start. I suppose we could at least have diff-start default to that. I\u0027ll give that some thought separately.","commit_id":"3d71ba11ee17af237f73fc97d3d3517812861b5a"}]}