Problem handling future timestamps?

[Wayne Davison]

On Thu, Apr 19, 2007 at 07:19:00PM +0100, Jon Burgess wrote:
> The particular problem I see where the timestamps 1940 != 2076 is due to
> a 64 bit time_t. On such platforms, increasing the modtime protocol
> entity from 4 to 8 bytes is sufficient to fix this.

Good point.  I'm going to include the ability to transfer a > 4-byte
time_t value for protocol 30.  We'll need to consider what should be
done if a value is too large for the destination system to handle.

..wayne..