)]}'
{"tests/test_bin.py":[{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"5a4601f75a899dbffacffdb99b980c6b519ad760","unresolved":false,"context_lines":[{"line_number":25,"context_line":"    def test_dbsync_run(self):"},{"line_number":26,"context_line":"        ret \u003d os.spawnlp(os.P_WAIT, \"python\","},{"line_number":27,"context_line":"                         \"python\", \"../bin/ceilometer-dbsync\","},{"line_number":28,"context_line":"                         \"--database_connection\u003dlog://localhost\")"},{"line_number":29,"context_line":"        self.assertEqual(ret, 0)"}],"source_content_type":"text/x-python","patch_set":3,"id":"AAAAM3%2F%2F%2Fl4%3D","line":28,"updated":"2012-12-17 11:29:32.000000000","message":"I think the openstack convention is to use the newer subprocess.Popen instead of the older os.spawn* library.\n\nApart from the consistency argument, Popen would allow you to drop the explicit reference spawning the python interpreter (and instead rely on the #! directive in bin/ceilometer-dbsync).","commit_id":"903200d247c051b307ef0d53ac0a13c19fc30976"}]}