and CURLOPT_CONNECTTIMEOUT_MS that, as their names should hint, do the
  timeouts with millisecond resolution instead. The only restriction to that
  is the alarm() (sometimes) used to abort name resolves as that uses full
  seconds. I fixed the FTP response timeout part of the patch.

  Internally we now count and keep the timeouts in milliseconds but it also
  means we multiply set timeouts with 1000. The effect of this is that no
  timeout can be set to more than 2^31 milliseconds (on 32 bit systems), which
  equals 24.86 days.  We probably couldn't before either since the code did
  *1000 on the timeout values on several places already.

So cookie expiration date is lowered to expire at most in 2035.

  fail since they used "1 feb 2007"...

- Manfred Schwarb reported that socks5 support was broken and help us pinpoint
  the problem. The code now tries harder to use httproxy and proxy where
  apppropriate, as not all proxies are HTTP...

  the problem. The code now tries harder to use httproxy and proxy where
  apppropriate, as not all proxies are HTTP...

upcoming libgcrypt-capable libssh2-versions

jar has died and we now instead point out our own version of that

Next time in 2038 :-)

open file descriptors is greater than FD_SETSIZE minus SAFETY_MARGIN,
also skip the test if any of the open file descriptors has a number
greater than FD_SETSIZE minus SAFETY_MARGIN.

<sys/time.h>.

of poll() and select() happens to be bound by FD_SETSIZE