Hello,

On Sat, Oct 26, 2002 at 08:29:27PM -0700, Tad Hunt wrote:
> [...] 
> Furthermore, it appears as though bringing a new interface up, or
> modifying the MTU of an existing interface doesn't cause any existing
> routes to be updated with the new information.

The explanation for this design is often cited as follows:

- the kernel is very simple-minded about routing.
- any higher order algorithm has to be implemented by userland programs
  (that is, by routing daemons).

So yes, in your second example, the higher order routing daemon (in this 
case, you yourself) would have to update the route. Or create it in the
correct order.

It can be argued that the _route(8) command_ should be responsible for this,
at least given a "-do-all-necessary-adjustments" switch. Any volunteer?

	-is
-- 
seal your e-mail: http://www.gnupg.org/