rja%batcomfs@Sun.COM (Robert Allen) (04/18/89)
I'm working on adding some 1984 X.25 features to a 1980 X.25 implementation
and I have a question about the facilites required by DDN X.25, as
specified in the DDN X.25 spec. in Vol. 1 of the DDN Protocol Handbook.
In section 2.1.2, it is noted that two facilities, type-of-service and
call precedence, should be included *after* all CCITT X.25 facilities and
must be preceded by a two byte facility marker, where each byte is of
zero value.
The type and location of the facility marker is counter to what is
documented in DIS 8208 "Info. Processing Systems - X.25 packet level
protocol for data terminal equipment", which states that:
(1) all CCITT specified facilities must come *after* "non-X.25
facilities supported by the network...of the calling DTE...",
which is what the DDN facilities are based on the facilities marker
used to denote them.
(2) that the facilites marker for X.25 specific facilities (which
happens to be two zero bytes) must come BEFORE the CCITT facilites
marker,
If, as the DDN X.25 spec states, the DDN X.25 facilities must come after
the CCITT facilities, *and* must use two zero bytes as the facilities
marker of two zero bytes, then the two protocols, DDN X.25 and 1984 X.25,
seem to not be compatible. Is this correct? If not, what am I mis-
understanding here?
Thanks in advance,
Robert Allen "I'm the NRA" (and soon the ACLU too.)
rallen@sun.com "Just say NO to the loss all our civil liberties."
Disclaimer: The opinions expressed are mine [so keep your hands off :-) ]malis@BBN.COM (Andy Malis) (04/18/89)
Robert,
I'm not familiar with DIS 8208. However, let me quote from the
CCITT Red Book (1984), Rec. X.25, section 7.1, p.209:
Facility/registration markers, consisting of a single octet
pair, are used to separate requests for X.25 facilities ...
from other categories as defined above [non-X.25 facilities
which may be provided by some networks] ...
The first octet of the marker is a facility/registration
code field and is set to zero. The second octet is ... set
to zero when the marker precedes requests for:
- non-X.25 facilities provided by the network in case of
intranetwork calls.
...
Facility/registration codes for X.25 facilities and for
the other categories of facilities may be simultaneously
present. However, requests for X.25 facilities must
precede the other requests.
So, you must first have the CCITT-defined X.25 facilities, then
the marker, and then the non-X.25 facilities provided by the
network.
Cheers,
Andy Malis
BBNCC PSN Developmentbarns@GATEWAY.MITRE.ORG (Bill Barns) (04/18/89)
I think you have become lost in CCITT/ISOspeak. This is no reflection on your intelligence or diligence; their nomenclature is considerably less than lucid. I'm not sure where you got your #1 from in 8208, but I'm speculating that you are looking near the end of clause 15.1. The statement actually made there is that "requests for X.25 facilities must precede the other requests and requests for CCITT-Specified DTE facilities must follow the other requests." In both 8208 and 1984 X.25, X.25 facilities and CCITT-Specified DTE facilities are two disjoint subsets of the facilities discussed in the documents. By X.25 facilities, they mean those discussed in clauses 6 and 7 of X.25(1984), which correspond to clause 13 of 8208. By CCITT-Specified DTE facilities, they mean those discussed in Annex G of X.25(1984), which corresponds to clause 14 of 8208. So the "non-X.25 facilities" (which as you correctly deduced, is what the DDN Type of Service and Precedence facilities are) are requested with "other requests". Translation: The X.25 facilities come first, the non-X.25 facilities (such as DDN's) come second, and the CCITT-Specified DTE facilities come third. Bear in mind that when the DDN X.25 spec in the Handbook was written, X.25(1984) wasn't out yet. Before X.25(1984), there was no such thing as a "CCITT-Specified DTE facility" and thus no reason to clarify where the DDN non-X.25 facilities belong in relation to them. So, when the DDN spec refers to "all CCITT X.25 facilities", it's unambiguous in the author's frame of reference. And strictly speaking, it isn't wrong now, if you parse it as (all (CCITT (X.25 facilities))) rather than as (all ((CCITT X.25) facilities)) since the CCITT-Specified DTE facilities are not X.25 facilities even though CCITT specifies them in Annex G of Rec X.25 and Annexes are considered an integral part of Recommendations. Sigh. (Fine print: CCITT specifies the coding of the CCITT-Specified DTE Facilities, but does not specify the procedures for using them, in the CCITT sense of the word "procedures". That issue is considered to be in the realm of "international user organizations" by which they mean ISO. Likewise, they don't tell you how the two types of "other requests" are positioned in relation to each other, only where they all are in relation to the facilities which are in their domain to specify. So the relative position of calling-DTE-specific-non-X.25-facility-requests and called-<what I just said> is a local network matter. You have to ponder this for a bit to realize that an internetwork call between two networks with conflicting rules doesn't make the conflict visible to either network!) The DDN doesn't specifically claim to support passing CCITT-Specified DTE facilities at this time, and I have no idea what would happen if you tried to do it. If it doesn't work now, it will probably be added on the next pass. Does it all make sense now? Hmm, let me rephrase that. Does it all seem consistent now? Bill Barns / MITRE-Washington / barns@gateway.mitre.org