)]}'
{
  "commit": "1ebafb61a4a2059fdd50aeefd27780d6358dd488",
  "tree": "5172fa876077c854d080ba4b2d639b1933ce932a",
  "parents": [
    "ac278ea6b770fd0bf74b3e718c2e6a4d84fcba3a"
  ],
  "author": {
    "name": "Paul Jakma",
    "email": "paul.jakma@hpe.com",
    "time": "Wed Apr 27 09:55:21 2016 +0100"
  },
  "committer": {
    "name": "Paul Jakma",
    "email": "paul@quagga.net",
    "time": "Fri Jan 27 14:53:31 2017 +0000"
  },
  "message": "bgpd: collision detection assumes \u0027new\u0027 peer is the inbound connection\n\n* bgp_packet.c: (bgp_collision_detect) for a long time, this has assumed\n  the \u0027new\u0027 peer argument on which an OPEN has just been received must be\n  an \u0027inbound\u0027 connection, and the looked up \u0027peer\u0027 the outbound. However,\n  this doesn\u0027t seem a robust assumption. It seems possible it could be the\n  other way around.\n\n  The consequences are that collision detection could behave inconsistently\n  with other implementations, and result in both sides closing the same\n  connection.\n\n  Fix to follow the RFC.\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "16bc4457fc674ed168d8557e562e455f61b44545",
      "old_mode": 33188,
      "old_path": "bgpd/bgp_packet.c",
      "new_id": "4ef470d2b7a037aa1791720e9853f24905135ee5",
      "new_mode": 33188,
      "new_path": "bgpd/bgp_packet.c"
    }
  ]
}