)]}'
{
  "log": [
    {
      "commit": "f8a2243ebeb3ce60e412cd7b71b31c7b31313d58",
      "tree": "c6c9bd55b48e97caebf60b5875544cdcce7f5755",
      "parents": [
        "1399759118f50071d499f88b18828e36aa2bad00"
      ],
      "author": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Tue Sep 22 12:34:13 2015 -0700"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Tue Sep 22 12:34:13 2015 -0700"
      },
      "message": "\u003cnetinet/udp.h\u003e should include \u003clinux/udp.h\u003e.\n\nThe comment about \"other stuff\" referred to pre-uapi headers. Everything\nin the current \u003clinux/udp.h\u003e should be exposed to userspace. The only\nproblem is that BSD and Linux use different names for the members of\nstruct udphdr. We can move the Linux udphdr out of the way and use an\nanonymous union to get the best of both worlds. (Though unfortunately\nthis means that code that includes \u003clinux/udp.h\u003e directly instead of\nusing \u003cnetinet/udp.h\u003e now won\u0027t have any definition of struct udphdr.\nWe\u0027ve taken the stance in the past that you shouldn\u0027t include a linux/\nheader if there\u0027s a standard equivalent --- you should rely on us\ntransitively including it for you.)\n\nChange-Id: Ie625892441b0edd8df3b76d3fcf2cbe299077bc4\n"
    }
  ]
}