**********************************************************************
FTSC FIDONET TECHNICAL STANDARDS COMMITTEE
**********************************************************************
Publication: FSP-1005
Revision: 6
Title: Zone 2 nodelist flags
Author: Frank Ellermann, 2:240/5815.1
Revision Date: 27 November 1997
Expiry Date: 27 November 1999
---------------------------------------------------------------------
Contents:
1. Introduction
2. FTS-0005 flags
3. User flags
4. Approved zone 2 user flags
5. Flag implications
6. Invalid combinations
7. Baud rate field
8. Thanks to...
---------------------------------------------------------------------
1. Introduction
---------------
This document informs about known differences of FidoNet zone 2
nodelist flags from FTS-0005.003. The ultimate sources for these
informations are the current Z2 nodelist epilogue and the setup
for flag corrections at Z2C, but it may be difficult to get these
sources for readers in other zones.
| All changes since version 5 are marked by a bar at the left edge.
It is (again) possible to list V32b and V42b in zone 2, upper case
V32B or V42B is not more enforced. Currently new flags needed for
IP-connectivity are under test in zone 2 (only internally), e.g.
-> VM VModem, default port 3141, dummy country code 000-
-> IFC IFCico, default port 60179, dummy country code 000-
-> BND BinkP, default port 24544, dummy country code 000-
-> IP general IP connectivity, dummy country code 000-
-> TELN Telnet dummy country code 000-
2. FTS-0005 flags
-----------------
The following flags are used as specified in FTS-0005.003:
CM Continuous Mail, node accepts mail 24 hours a day
MO Mailer Only (no BBS)
LO Listed Only, node accepts calls only from listed
node numbers in the current FidoNet nodelist
-> V21 ITU-T V21 300 bps full duplex (obsolete)
V22 ITU-T V22 1200 bps full duplex (obsolescent)
| In zone 2 the value 1200 in the baud rate field implies V22. Only
| two nodes not supporting at least V22bis, ISDN, or IP still exist
| in the zone 2 segment. Flag V22 is almost obsolete, and V21 is
| already removed in Z2. Both flags should be dropped from the next
| FTS-0005 version.
V29 ITU-T V29 9600 bps half duplex (obsolescent)
-> V33 ITU-T V33 14400 bps half duplex (obsolete)
V33 cannot be used in connecting FidoNet nodes over public dial-up
lines and is most probably a historical error in FTS-0005. A very
similar argument is applicable on V29, all nodes flying this flag
support at least V32. Today only one node in Z2 still flies V29,
and V33 is already removed in Z2. Both flags should be dropped in
the next FTS-0005 version.
V32 ITU-T V32 9600 bps full duplex
V32b ITU-T V32bis 14400 bps full duplex (implies V32)
| V34 ITU-T V34 28800 bps full duplex (33600 bps option)
FTS-0005 specifies V32b and V42b (capital V and small b), current
nodelist practice in FidoNet shows all combinations of small and
capital letters for flags. This was no problem before FSC-0062
introduced case-sensitive flags. The best solution is to stick to
the current practice and treat all old flags as case-insensitive.
H96 Hayes V9600
HST USR Courier HST up to 9600 (implies MNP)
H14 USR Courier HST up to 14400 (implies HST)
-> H16 USR Courier HST up to 16800 (implies H14 and V42b)
MAX Microcom AX/96xx series (almost obsolete)
PEP Packet Ensemble Protocol
CSP Compucom Speedmodem
-> ZYX Zyxel series 16800 bps (implies V32b and V42b)
-> V32T V.32 Terbo 19200 bps (implies V32b)
VFC V.Fast Class 28800 bps (should imply V32b and V42b)
If a flag directly or indirectly implies other flags, then these
other flags are not shown in a nodelist entry, because this would
be redundant. Unfortunately the rules for redundancies in zone 2
and FTS-0005 are different. Zone 2 continued to avoid redundancy
with most "new" flags, but FTS-0005.003 specified no redundancies
for "new" flags like ZYX, H16, V32T, or VFC. "New" flags in this
context are flags approved by FidoNet International Coordinators
since 1989, when FTS-0005.TXT, the predecesssor of FTS-0005.003,
was published.
For details see the chapter "implications" below, for now only
note, that in zone 2 H16 implies V42b, ZYX implies V32b and V42b,
and V32T implies V32b.
Zone 1 and zone 2 have introduced a user flag Z19 approved by the
corresponding Zone Coordinator. User flags are discussed later,
for now only note, that in zone 2 ZYX is specified as Zyxel 16k8,
while FTS-0005.003 not knowing Z19 specifies ZYX as generic flag
for all Zyxel protocol speeds.
Today only one node in FidoNet still really flies MAX, this flag
is obsolete and should be dropped from FTS-0005. The flags CSP
(7 nodes worldwide) and H96 should be marked as obsolescent.
| MNP Microcom Networking Protocol 2-4 error correction
| V42 ITU-T LAP-M error correction w/ fallback to MNP 2-4
| V42b ITU-T V.42bis BTLZ data compression over V.42 LAP-M
The next version of FTS-0005 should adopt the better V42b and
MNP definitions of the zone 3 nodelist epilogue. FTS-0005.003
specifies an implication of V42 by V42b, but the exact meaning of
the flag MNP is unclear. Most probably this flag was meant to
| indicate support of MNP 2-4, and in this sense V42 implies MNP.
| Note the difference between the flag V42b (implying V42) and the
| standard V.42bis (not necessarily based on LAP-M as data link
| layer protocol), without this difference the flag V42b would be
| ambiguous for combined modem and ISDN node entries.
MN No compression supported in insecure inbound
XA Bark and WaZOO file/update requests
XB Bark file/update requests, WaZOO file requests
XC Bark file requests, WaZOO file/update requests
XP Bark file/update requests
XR Bark and WaZOO file requests
XW WaZOO file requests
XX WaZOO file/update requests
These flags are equivalent in FTS-0005 and in the zone 2 segment.
Gx..x Gateway to domain 'x..x'
Valid values for this flag are assigned by the Fido International
Coordinator, FTS-0005.003 explicitly mentions GUUCP. In zone 2
only GUUCP gateways are flagged.
#01 Zone 5 mail hour (01:00 - 02:00 UTC) w/ Bell 212A
#02 Zone 2 mail hour (02:30 - 03:30 UTC) w/ Bell 212A
-> #08 Zone 4 mail hour (08:00 - 09:00 UTC) w/ Bell 212A
#09 Zone 1 mail hour (09:00 - 10:00 UTC) w/ Bell 212A
#18 Zone 3 mail hour (18:00 - 19:00 UTC) w/ Bell 212A
#20 Zone 6 mail hour (20:00 - 21:00 UTC) w/ Bell 212A
The variants !01, !02, !08, !09, !18, and !20 indicate missing
Bell 212A support. In zone 2 #02 or !02 would be obviously
redundant.
Today less than four 1200 modems (V22 or Bell 212A) are listed.
A future version of FTS-0005 should drop !mn variants together
with V21 and V22 flags.
Further most non-CM systems flagging #mn or !mn today probably
want to show additional online times instead of additional mail
hours. As soon as FSC-0062 flags have been approved by the IC
or adopted as FTS by the FTSC, the following version of FTS-0005
should mark #mn as obsolescent and recommend the more flexible
FSC-0062 flags (see below).
3. User flags
-------------
An example for one of several problems in zone 2 with user flags:
...,U,Z19,V110H,V120L,V120H,X75,ENC,NEC
These flags indicate a modern Zyxel ISDN-modem and two additional
user flags ENC and NEC. This possible user flags string contains
34 characters, but at most 32 characters are allowed in FTS-0005.
...,U,Z19,V110L,V110H,X75,ISDNA,ISDNB,ISDNC
During the period for the replacement of old by new ISDN flags
(several months !) many nodes listed both old and new flags for
maximal compatibility, and no problems with nodelist compilers
or mailers caused by too long user flags strings were reported.
Therefore the length limit in FTS-0005 is probably unnecessary
and at least inconsequent: Other nodelist fields like the system
name are unlimited, so why only restrict the user flags string ?
To help developpers an upper limit of e.g. 255 characters for a
nodelist line and 63 characters for fields 3 to 6 would be more
useful.
The next problem with user flag strings as specified in FTS-0005
is their introduction by the letter U with no comma following:
Nodelist compilers could parse ...,UISDN,USR in user flags ISDN
and USR. But USR cannot be approved as "real" flag, because the
combination ...,USR,UISDN would then be parsed in SR and UISDN.
Other side effects of the FTS-0005 specification are additional
difficulties in finding flags. Almost all flags are separated
by a comma, only the first user flag can be an exception to this
simple rule. If the order of user flags has no meaning, then...
...,UV120L,V120H
...,UV120H,V120L
... are equivalent. A "simple" solution of this problem could be
to treat UV120L as synonym for V120L, and UV120H as synonym for
V120H. Similar problems show up, if user flags are counted, etc.
Obviously a nodelist compiler looking for user flags has always
to consider the case "user flag separated by comma". In zone 2
this idea was simply extended to the first user flag:
All flags are separated by commas. Flags not yet approved by the
International Coordinator or the FTSC (i.e. user flags only used
experimentally or locally) are separated by a new pseudo flag U.
-> U pseudo flag to the left of at least one user flag
All flags following this pseudo flag U are user flags, all flags
before this pseudo flag are "real" flags specified in FTS-0005 or
approved by the International Coordinator.
Because this definition should be compatible with any reasonable
software implementation based on FTS-0005.003, and simplifies the
handling of user flags significantly, a future FTS-0005 version
will hopefully adopt it.
4. Approved zone 2 user flags
-----------------------------
In zone 2 user flags have to be approved by the Zone Coordinator.
Currently the following zone 2 user flags exist:
-> V110L ITU-T V.110 19k2 async 'Low' (former ISDNA)
-> V110H ITU-T V.110 38k4 async 'High' (former ISDNB)
-> V120L ITU-T V.120 56k6 async, N1 = 259, W = 7, modulo 8
-> V120H ITU-T V.120 64k async, N1 = 259, W = 7, modulo 8
-> X75 ITU-T X.75 SLP (single link procedure),
64kbit/s B channel; layer 2 max. framesize N1 = 2048,
window size W = 2, frame numbering modulo 8;
layer 3 transparent (no packet layer)
-> ISDN Other configuration, used only if none of above fits
These ISDN flags follow the specification in FSC-0091.
-> Tyz Online time flags as specified in FSC-0062
The flag Tyz is used by non-CM nodes online not only during ZMH,
y is a letter indicating the start and z a letter indicating the
end of the online period as defined below (times in UTC):
A 0:00, a 0:30, B 1:00, b 1:30, C 2:00, c 2:30,
D 3:00, d 3:30, E 4:00, e 4:30, F 5:00, f 5:30,
G 6:00, g 6:30, H 7:00, h 7:30, I 8:00, i 8:30,
J 9:00, j 9:30, K 10:00, k 10:30, L 11:00, l 11:30,
M 12:00, m 12:30, N 13:00, n 13:30, O 14:00, o 14:30,
P 15:00, p 15:30, Q 16:00, q 16:30, R 17:00, r 17:30,
S 18:00, s 18:30, T 19:00, t 19:30, U 20:00, u 20:30,
V 21:00, v 21:30, W 20:00, w 20:30, X 23:00, x 23:30.
For example TuB shows an online period from 20:30 until 1:00 UTC.
-> Z19 Zyxel series 19200 bps (implies ZYX)
-> X2C x2 client w/ 56000 bps (should imply V34 and V42b)
-> X2S x2 server w/ 64000 bps (should imply V34 and V42b)
-> K12 Systems offering all educational K12-conferences
-> ENC The node accepts inbound encrypted mail
-> NC Network Coordinator (only if the NC is not the host)
-> NEC Net Echomail Coordinator (at most one per net)
-> REC Region Echomail Coordinator (at most one per region)
Redundant AKAs used to indicate echomail coordination in zone 2
are no longer permitted. One *EC flag is valid for all AKAs of
a given sysop.
5. Flag implications
--------------------
Flag implications directly or indirectly specified in FTS-0005:
HST => MNP
H14 => MNP HST
H16 => MNP HST H14
V42b => V42 (MNP ?)
V32b => V32
Flag implications specified in the zone 2 nodelist epilogue:
HST => MNP
H14 => HST MNP
-> H16 => V42 MNP V42b H14 HST
-> V42b => V42 MNP
-> ZYX => V42 MNP V42b V32 V32b
-> Z19 => V42 MNP V42b V32 V32b ZYX
V32b => V32
-> V32T => V32 V32b
-> V110L => ISDN
-> V110H => ISDN
-> V120L => ISDN
-> V120H => ISDN
-> X75 => ISDN
The latter ISDN flag redundancies are a consequence of FSC-0091.
Maybe some of the following implications could be added in zone 2:
VFC => V32 V32b MNP V42 V42b
X2C => V34 MNP V42 V42b
X2S => V34 MNP V42 V42b
Flag implications (i.e. not listing redundant flags) have several
advantages: Some old nodelist tools are unable to handle too long
lines. Old flags like HST, MNP, V42, or V32 vanish automatically,
if they are implied by H16, V42b, V32b, or better. Redundancies
defined globally for the whole nodelist help to avoid flag errors.
6. Invalid combinations
-----------------------
All file request flags exclude each other (at most 1 is possible):
XA, XB, XC, XP, XR, XW, and XX. For flag checkers only supporting
implications a good approximation based on FTS-0005 definitions is
| XA => XW XR XP XB XC XX,
| XB => XW XR XP,
| XC => XW XR XX,
| XR => XW,
| XX => XW.
Further X2C cannot be combined with X2S, and FSC-62 Tyz-flags are
not possible with CM. Also Tyz with y = z is of course incorrect.
Some modem protocols are "proprietary" in a sense, that all today
known modems can fly at most one of the corresponding modem flags:
MAX, CSP, H96, PEP, HST, H14, H16, ZYX, and Z19.
A few "old" modem protocol flags are known to be invalid if used
together with "new" protocol flags, i.e. each "old" flag excludes
all "new" flags and vice versa:
"Old" in this sense are MAX, CSP, H96, HST, H14, V32, and PEP.
"New" in this sense are X2S, X2C, V34, VFC, V32T, and H16.
For Z2 add ZYX as "old" and Z19 as "new". A simple REXX script to
test some known inconsistencies is available as NLSCHECK.REX at
the site of the author. While erroneously listing redundant flags
causes no harm, other errors like combining V34 with HST or Z19
with H16 indicate serious problems, which can result in connection
failures or other damage.
7. Baud rate field
------------------
The baud rate field 7 in the nodelist as specified in FTS-0005 is
nearly useless today: Except from a few remaining 1200 and 2400
nodes almost all nodelist entries show either 9600 for all modem
protocols better than V22bis or 300 for ISDN (or IP) only nodes.
No more V21 or Bell 103 modems are listed for more than 2 years.
The baud rate values 19200 and 38400 specified in FTS-0005.003
have not been used in the FidoNet nodelist. So all a reasonable
nodelist compiler can do today, is treat 300 as indicator for
ISDN or IP only, and treat unknown or missing values in field 7
like 9600.
A new meaning for field 7 as speed field could be really useful.
An example is ZYX, if we would have 16800, 19200, 28800, and 33600
as speed values, then their combination with ZYX is all we need
technically, Z19 would be unnecessary. Another example is HST,
flags H14 and H16 are unnecessary, if HST is combined with 9600,
14400, 16800, 28800, or better. Variants of PEP could be shown in
the speed field without new flags. "Enhanced V32.terbo" could be
shown by 21600.
Most important: V34 may have the famous bug not allowing connects
from new "V34+", unless the caller disabled symbol rate 3429. If
"V34+" is indicated by speed 33600 or better, then an appropriate
setup for all kinds of V34 connects is possible.
A future version of FTS-0005 hopefully allows the following speed
values in field 7:
300 reserved for ISDN or IP only (for historical reasons)
1200 obsolete (either V.22 in Z2 / Z3, or Bell 212A in Z1)
2400 implies V22bis, qualifies as least common denominator
9600 default, used with PEP, V32, HST, H96, (CSP), (MAX)
12000 rare variant of V32
14400 used with V32b or HST (obsoleting H14)
16800 used with ZYX or HST (obsoleting H16)
19200 used with V32T or ZYX (obsoleting Z19)
21600 rare variant of V32T (no "H21" needed)
28800 used with VFC or V34
33600 used with V34 (no V34+ or V34b needed)
| 56000 used with X2C, X2S, or V.PCM
Allowing more than 12 speed values or allowing speed values above
64000 could break existing software (MakeNL, V7). Therefore the
next step in FidoNet could be, to add 12000, 14400, 16800, 19200,
21600, 28800, 33600, and 56000, where 19200 is already specified
in FTS-5 since 1989.
8. Thanks to...
---------------
Ben Baker St. Louis nodelist format
Rick Moore FTS-0005.TXT
David Nugent FTS-0005.003 and NLTOOLS
Jonny Bergdahl ERRFLAGS 2.6
Ward Dossche Zone 2 nodelist epilogue
David J. Thomas FSC-0062.003 (FRL-0062)
Jan Ceuleers FSC-0075.001 (FRL-0075)
Arjen Lentz FSC-0091.001 (FRL-0091)
Leonard Erickson CHECKNL 2.14 and many discussions in NET_DEV
Jim Barchuk LNDL 2.7
Marius Ellen FASTV7 2.04
| Jan Vermeulen, Ian Smith, Gisbert Rudolph, Carlos Fernandez Sanz,
| Tom Schlangen, Craig Ford, Pedro Lima, and many others...
**********************************************************************
Go Back