On Mon, Feb 02, 2004 at 11:39:30AM +0100, Wijnen, Bert (Bert) wrote: > > > > I think 0 makes sense for the default route as explained by Mike. > > Otherwise, if the idea is indeed to exclude 0, then the range should > > be (1..4294967295) - this makes it quite clear that just 0 was removed > > from the set of possible values. > > > I understand that zero makes sense. I did not say we should remove it. > I said that by specifying the range (and including zero) that that way > we are explicit to state that we do want it included.
I am not sure I like the direction to specify ranges which just repeat what is there anyway. See below... > By the way I like Mike's proposal to make it (0..2040), which would > line up with the max size of an InetAddress (255 octets). This makes indeed sense. One can ask whether this range should not be added to the InetPrefixLength TC itself so that the range is already bound to the TC. Of course, applying to your logic above, you would repeat the range (0..2040) anyway just to be explicit that zero is included... So what do we do? a) Add the range (0..2040) to the InetPrefixLength TC? (Now is the right time to do this.) b) Add the range (0..2040) just to the objects in question that use the InetPrefixLength TC? c) Do both? I think I prefer a) at the moment. /js -- Juergen Schoenwaelder International University Bremen <http://www.eecs.iu-bremen.de/> P.O. Box 750 561, 28725 Bremen, Germany -------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
