Hi, Fred (et al.),
It might be useful to be clear whether this option MUST NOT be used on atomic
datagrams (i.e., where IPv4 DF==1 or when not source fragmented) and that it
cannot be used for purposes other than reassembly (as the regular ID is per RFC
6484).
Joe
—
Dr. Joe Touch, temporal
Hi, Fred (et al.),
It might be useful to be clear whether this option MUST NOT be used on atomic
datagrams (i.e., where IPv4 DF==1 or when not source fragmented) and that it
cannot be used for purposes other than reassembly (as the regular ID is per RFC
6484).
Joe
—
Dr. Joe Touch, temporal
On Sat, 29 Jul 2023 at 16:43, to...@strayalpha.com
wrote:
> Hi, Fred (et al.),
>
> It might be useful to be clear whether this option MUST NOT be used on
> atomic datagrams (i.e., where IPv4 DF==1 or when not source fragmented) and
> that it cannot be used for purposes other than reassembly (as t
I would agree it must not be there when DF is 1. When DF is zero, there is
always the possibility that a forwarding router will fragment the packet. From
a practical standpoint, one wants the originator of the packet to provide the
option whenever sending to a different subnet.
I would also s
[Resending from Moderators account]
Hi all
This draft (draft-templin-intarea-ipid-ext-00) is being discussed across
three different lists - intarea, ipv6 and the IETF list. As the latter
list is for general topics only [1], I am removing it from the thread.
Please continue to discuss on the i
