In some email I received from Charles M. Hannum, sie wrote:
> 
> 
> As I pointed out, and Steiner verified, there are existing hardware
> implementation that do *not* verify the checksum.  So arguing that RFC
> 1812 (which isn't even a standard) requires it is a bit silly.

Do you have any objections to a sysctl to toggle this (small performance
hit for the check aside) behaviour ?

Or on the other hand, how many people (besides Perry) would actually
enable checksumming on hosts doing gatewaying for packets being forwarded
given the pros and cons as discussed to date ?

Darren