minor spec inconsistancies
Robert Woody Woodburn <woody@sparta.com> Wed, 13 May 1992 20:19 UTC
Received: from nri.nri.reston.va.us by ietf.NRI.Reston.VA.US id aa03764;
13 May 92 16:19 EDT
Received: from nri.reston.va.us by NRI.Reston.VA.US id aa05891;
13 May 92 16:25 EDT
Received: from PARK-STREET.BBN.COM by NRI.Reston.VA.US id aa05887;
13 May 92 16:25 EDT
Received: from park-street by PARK-STREET.bbn.COM id aa28547;
13 May 92 16:04 EDT
Received: from BBN.COM by PARK-STREET.BBN.COM id aa28541; 13 May 92 16:01 EDT
Received: from SPARTA.COM by BBN.COM id aa15624; 13 May 92 15:59 EDT
Received: from worf.SPARTA.COM by sparta.com (5.65/1.34)
id AA11580; Wed, 13 May 92 16:04:10 -0400
Received: by worf.sparta.com (5.65/1.34)
id AA02563; Wed, 13 May 92 16:04:05 -0400
Date: Wed, 13 May 92 16:04:05 -0400
From: Robert Woody Woodburn <woody@sparta.com>
Message-Id: <9205132004.AA02563@worf.sparta.com>
To: idpr-wg@bbn.com
Subject: minor spec inconsistancies
Martha,
I was going back through some protocol specific numbers in the spec and
found a few inconsistancies. I also found some other things that I'll
mention first:
1. I thought we had agreed to move int/auth to the end of the CMTP
header. It is still in a fixed place. It doesn't matter tons,
but I've already coded a bunch of routines the one way. It can
be changed later (not now), but I wanted to see if I'd forgotten
something. I can see some rational for always having the authentication
in the same place, but I also see reasons not to.
2. None of the IDPR protocols document the use of the INFORM field. It
really was used as a hack for the previous prototype, but it is
no longer necessary. Since none of the protocols document it, I think
it should be removed.
3. Addition of another NAK error type: I think it is useful to have
another error type so that IDPR protocols can report reasons they
reject an IDPR message. There currently is no way to do this, and
there really is no need for a per protocol error message. The
IDPR protocol would place a reason for the nak in the err_info field.
These would be protocol specific based upon the DPR and DMS fields in
the CMTP header.
4. Now for the inconsistancies: In most places you start at 0, other
places you don't. Some messages have values that conflict with others.
I have taken to the convention of counting from 0. Below, XXX notes a
where a section differs from the spec. Look for "spec says" for the
specific differences.
/* CMTP message types as per the March 1992 spec*/
#define CMTP_DATA 0
#define CMTP_ACK 1
#define CMTP_NAK 2
#define CMTP_MSGTYPE_MIN CMTP_DATA
#define CMTP_MSGTYPE_MAX CMTP_NAK
/* XXXX Differs from the spec XXXX */
/* I bet you left a whole for 1 = encap, but that went away since */
/* encapsulation is a separate IP protocol. */
#define CMTP_PROTO_VGP 0 /* spec says 2 */
#define CMTP_PROTO_RID 1 /* spec says 3 */
#define CMTP_PROTO_RSQP 2 /* spec says 4 */
#define CMTP_PROTO_PCP 3 /* spec says 5 */
/* Implementation specific stuff */
#define CMTP_PROTO_MIN CMTP_PROTO_VGP
#define CMTP_PROTO_MAX CMTP_PROTO_PCP
/* VGP message defs as of Mar 1992 IDPR spec*/
#define IDPR_VGP_UD 0
#define IDPR_VGP_PG_CON 1
#define IDPR_VGP_PG_PLCY 2
#define IDPR_VGP_VG_CON 3
#define IDPR_VGP_VG_PLCY 4
#define IDPR_VGP_MIN IDPR_VGP_UD
#define IDPR_VGP_MAX IDPR_VGP_VG_PLCY
/* RID message defs as of March 1992 IDPR spec */
#define IDPR_RID_CFG 0
#define IDPR_RID_DYN 1
#define IDPR_RID_MIN IDPR_RID_CFG
#define IDPR_RID_MAX IDPR_RID_DYN
/* RSQP Protocol message types as per March 1992 spec */
/* XXX These differ from the spec somewhat XXX */
#define IDPR_RSQP_RID_RQST 0 /* spec says 2 */
#define IDPR_RSQP_RT_RQST 1
#define IDPR_RSQP_RT_RESP 2
#define IDPR_RSQP_MIN IDPR_RSQP_RID_RQST
#define IDPR_RSQP_MAX IDPR_RSQP_RT_RESP
/* PCP Protocol message types as per March 1992 spec */
#define IDPR_PCP_SETUP 0
#define IDPR_PCP_ACCEPT 1
#define IDPR_PCP_REFUSE 2
#define IDPR_PCP_TEARDOWN 3
#define IDPR_PCP_ERROR 4
#define IDPR_PCP_REPAIR 5
/* Implementation specific extensions */
#define IDPR_PCP_SETUP_RQST 6
#define IDPR_PCP_TEARDOWN_RQST 7
#define IDPR_PCP_ERROR_RQST 8
#define IDPR_PCP_MIN IDPR_PCP_SETUP
#define IDPR_PCP_MAX IDPR_PCP_ERROR_RQST
wood y
- minor spec inconsistancies Robert Woody Woodburn