a while due to all the individual read() calls).

  been supported in Synchronet for quite some time, now.
* Force a wait of 1 second if there are no clients or servers connected to
  the IRCd, fixing the 100% CPU usage bug under Win32.  Bug reported by
  Lord Time on DOVE-Net.  Thanks!

error  - this is the wrong error value for Windows and it shouldn't be logged
using lprintf (to the Telnet log/window). Instead, this JS function will now
return null if/when select returns < 0 and the script can handle the
error condition and log or not log as it sees fit.

This should work but seems suspiciosnly like rev 1.91 which needed
modification to work on OSX.

even if the build system would like to use ncurses (ie: NetBSD)

regardless of the compiler.

  causing it to use 100% CPU until the first client connected.

  (randomly) swap the 'hops' and 'created' variables on the NICK line without
  warning.  The IRCd now tests for this.  This manifested itself with time_t
  integers appearing in the 'WHO' command output when a CR server was linked.
* WHOWAS output would always show the uplink server for a client who left the
  network from *behind* the uplink.  The correct server info is now shown on
  WHOWAS.

presses F1 for any reason.

restrictions (though the 'P' restriction should normally already be there).

no command is specified, executes SHELL

with SDL audio if we need to in the future.

This paves the way to removing the SDL from xpdev again (finally)
Which in turns paves the way for having SDL enabled in ciolib apps again.

going to DO with them, but here they are!