(sipp) DRAFT minutes of SIPP working group

[Jim DeMarco <[e-mail]>]

I have a _small_ concern about the following points in the SIPP spec
(on Page 24):

>   o  Unlike IPv4, when UDP are originated by a SIPP node, the UDP
>      checksum is not optional.  That is, whenever originating a UDP
>      packet, a SIPP node must compute a UDP checksum over the packet
>      and the pseudo-header, and, if that computation yields a result of
>      zero, it must be changed to hex FFFF for placement in the UDP
>      header.
>
>[...some lines deleted...]
>
>When the remote address is that of an IPv4-only system, a UDP Checksum
>field of zero in an incoming packet indicates the absence of a UDP
>checksum.  Such packets shall be accepted as is.  However, a SIPP system
>must always compute a valid (non-zero) checksum for outgoing UDP
>packets.
>
>[...some lines deleted...]
>
>Expires: January 17, 1995                                      [Page 24]

Although it is not stated _explicitly_, the implication here is that
a packet with a UDP Checksum field of zero SHOULD/MUST be *DISCARDED*
(excepting the case when the remote address is that of an IPv4-only
system).

If this is indeed the case, could this be stated _explicitly_,
please?  Otherwise *some* vendor(s) *will* cut corners in the
interest of performance (it has happened before) and everyone else
will have to accept those zero-checksum packets (those who _are_
"conservative in what they send" will still "be liberal in what they
accept").  An explicit statement that such packets MUST be dropped
will go far in nipping this in the bud.

That is, unless I am wrong in my interpretation and zero-checksum
UDP packets *SHOULD/MUST NOT* be discarded.

What _was_ the actual intent?

--Jim
 ------------------------------------------------------------------------------
IETF SIPP Working Group - Archives:  parcftp.xerox.com:/pub/sipp
Unsubscribe:	unsubscribe sipp		(as message body, not subject)
Direct all administrative requests to [e-mail]