[freespace2] some stuff...

[Steven Fuller]

On 31 Jul 2003, Taylor Richards wrote:

> One question while I'm going through old code, SOCKET is defined as an
> int but on windows it's unsigned.  I know that play between Linux and
> Windows works but could this cause a problem somewhere down the line?  I
> noticed this while cleaning up the compiler messages in the networking
> code.  Making it unsigned is obviously a problem since INVALID_SOCKET is
> -1 and that's more work to get around.  So basically I have a couple of
> compiler warning fixes that aren't in this patch since I don't know the
> best way to deal with it.

I don't think it's a problem in this case.  No arithmetic is performed
with the socket values and as long as sockets are compared against
INVALID_SOCKET and not less than zero, it should work fine as-is.


Steven Fuller