home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
HAM Radio 1
/
HamRadio.cdr
/
misc
/
rs150upd
/
rs14xupd.rme
< prev
next >
Wrap
Text File
|
1991-03-04
|
28KB
|
640 lines
This is a summary of ReadMe's from versions 1.43-1.48 extraneous
comments etc have been removed for brevity
PRMBS v1.44 - 07/28/90
CONFIG file changes - removal of a number of lines at the end. You should
be able to convert 1.43/1.42 files by simple removal. All lines are removed
from the end of CONFIG before the AUTO-EVENT stuff. All changes take place
from file spec section down. I have enclosed a sample CONFIG.CON only since
changes will be identical in all type of configs. Be sure to remove the entry
for the OLDMAIL.KLD file as well as the paremters down lower. Make sure you
read below about some additions that will most likely be needed to the
AUTO-EVENT section.
I have made some changes as well as fixes in V1.44
- the bug with more than 32 MBytes free producing disk full
- some guys complained of inadvertent crashes because of hitting keyboard
I added CTRL-L as a KB lock - if the KB lock is on only a CTRL-L
will allow keys to be hit, the ESC, the A-dir, the L, the H, etc
- the problem with the sysop placing a home-bbs in his user record to
force the insertion of a "Reply-to:" field in his messages causing
his inbound mail to be re-addressed and sent back out has been fixed
by a new parameter (see below the section on SETPARM) called HOLDSYSOPMAIL
this defaults to TRUE and will stop the re-addressing. If the sysop
is leaving his system on and wants his mail sent elsewhere, then he
adds the command SETPARM HOLDSYSOPMAIL OFF in his AUTO-EVENT or EVENT
files.
*** Please take note of the changes in comments in the files sections of the
CONFIG files. I had some misleading commenst in there and have rectified that
stuff. The PASSWORD file entry is a path and file spec - you may place it
anywhere you wish. It is not necessarily found in the SYS directory - I am
sorry for the confusion past documentation has caused on this matter. As
shwon in the included sample CONFIG the password file will be found in the
main BBS directory, NOT in the SYS directory.
The changes are as follows;
- lstmsg() procedure (E,L,K,R,FILE,FLOOD commands)
now has two additional '-' options
'-t' - says traffic only (like '-b' is bulls only)
'-f filename' specifies a file name for the 'file'
command. no filename specified defaults to PRMBS.EXP
- the OLDMAIL.KLD file spec is gone and both CONFIG options
about archiving are gone and the archiving in MAILWORK compress
is gone ....
to archive killed traffic only you add to your event file
before the COMPRESS command
file -t -k -f oldmail.kld all
this will archive all killed traffic to OLDMAIL.KLD
to archive everything leave out the '-t'
then run the COMPRESS command
*** the above is in keeping with the spirit of PRMBS in that rather than
have lots of special features we are using primitive existing commands to
achieve complex things - archiving is no more than file messages under
special circumstances sooooo ...
- all but two of the parameters that were in he CONFIG file that were
simple ON/OFF or simple numbers have been combined into a single command
SETPARMS which takes up to two arguments. Below is a sample for the
AUTO-EVENT section of your config to set these parameters. BY making them
command line parms thet can be changed by sysop or remote sysop without
having to reboot and can be modified at different time the day by EVENT
files.
M 0100 SETPARM MAXERROR 6
M 0100 SETPARM MODEMWAIT 900
M 0100 SETPARM PONG 4
M 0100 SETPARM AUTOFLOOD ON
M 0100 SETPARM BELL OFF
M 0100 SETPARM CHAT OFF
M 0100 SETPARM FORCETYPE ON
M 0100 SETPARM LOGGING ON
M 0100 SETPARM LOGMSGSONLY OFF
M 0100 SETPARM LOGLOCAL ON
M 0100 SETPARM NOBIDFL ON
M 0100 SETPARM OPENLO ON
To see the values of these parameters you may type SETPARM and the values
of all of the above will be ddisplayed. If you type SETPARM {parm-name} it
will give you the value of just that parameter. (for example SETPARM PONG)
. You only need to type as few characters as needed to make the par,-name
unique in the list above.
*** NOTE *** you may wish to leave all of the above parameter settings out
of the AUTO-EVENT section to start and birng up a console version and do a
SETPARM and see what the default settings are.
- you will see a new parameter up there it is FORCETYPE. If this is ON the
system behaves as it did in the past. If the user typed S or Sx (where 'x' is
other than a B,P, or T) the system will assign a type P or B depending on
whether the TO field is a valid CALL or not. If it is set to OFF a simple S
to send will produce a message with no type. *** WARNING *** if such a
message is forwarded off of the system, unless the adjacent system is a PRMBS
with this parm set OFF also or a CBBS (k3rli) system, the message will get
typed. This applies to MBL, RLI, REBBS,MSYS,and PRMBS (pre-1.44 and 1.44 with
FORCETYPE ON)
- a recent loss of an important mail serve system showed up the power of
the POLLF command, but brought to light a small deficiency that is even
though the two system that took up the task could do a POLLF for mail for the
downed system, RMAILS would go unexploded. The 'D' option in the sweeper
which -reruns flooding for bulls now tests the recipient of the message if it
is RMAIL it will do an RMAIL explosion on the spot regardless of the target,
the parent message will get its status changed to 'F' - notice this is a
manual command - I see too many problems with making it automated ... I gotta
work on that.
PRMBS v1.45 - 08/07/90
One small fix over v1.44 - I finally found why the system will often, BUT
NOT ALWAYS, crash after doing a compress. Safe to say (I know thats what I
said the last time I 'fixed' this problem) its fixed now I tested it under
the most extereme failure mode - once I understood exactly where it was dying
and its now - "harry-proofed"!!!
removed MODEMWAIT from SETPARM command - not needed any more
added MINDISKSIZE and REQFILMAX to SETPARM command
now there are no operational parameters in he CONFIG file, only
file and directory paths/names, BBS IDs and PORT parameters.
PRMBS/ROSERVER v1.46 09/20/90
CONFIG changes - usual version and date
MESSAGES.RS changes, add the followimg two messages; I suggest to place
them about 20 from the top:
154 OLD BULLETIN - HELD for review by SYSOP
155 Possible MESSAGE LOOPING - HELD for review by SYSOP
H_LIST.RS changes, to numerous to mention. If you have done extensive editing
on your file I have included a program called DIFF and included the H_LIST as
H_LIST.NEW. Type
DIFF H_LIST.RS H_LIST.NEW
the display should be self explanatory - it is a unix style diff, with
intelligent analysis of differences. If you have sorted your H_LIST, DIFF
won't be much help. The output of DIFF can be re-directed to a file for
examination or paged if its humongous.
DIFF H_LIST.RS H_LIST.NEW >filename (redirection)
DIFF H_LIST.RS H_LIST.NEW | MORE (screen paging)
FIXES:
1 - the lockup from v1.4x SEEMS to be fixed. After doing a COMPRESS an
L command would send it to never never land. To be honest I am not quite
sure exactly which of several tinkerings I did 'fixed' it. I have only
the vaguest idea of exactly what went wrong, though I know exactly
where it was spinning in a hard loop. I will be keeping an eye on this
for recurrance.
2 - if a space bar was hit during an H or M or L command when in
monitor mode the system would stay locked, I put a 2 minute timer in
the pause loop then it goes on about its business. Actually the timer
applies to any time the SPACEBAR or CTRL_S is used.
3 - system wasn't logging over the air forwarding of mail correctly, it
now is.
CHANGES/ENHANCEMENTS:
Changes to EX/KX - 'expired' bulletins commands, the presentation order
is from newest messages in file backwards, as it should have been, not
my goofy idea awhile back to strat at the front.
Some SETPARM additions
BULLEXPIRE (default = 0) numeric argument - if non-zero it specifies
an age of a bulletin in days. IF an incoming Bull is older than the
days specififed it stores as an 'H' status and is not flooded. A
message from MESSAGES.RS (#154) goes out just before the usual 'RCVD'
message to the sender.
LONGIHEADER (default ON) ON/OFF argument - toggles the inclusion/exclusion
of the "date: " and "Subject: " fields in the internal RFC-822 header
of locally generated headers. When OFF only the "To:", "From: ", "Message-ID:"
and "reply_to:" (when required) are included.
LONGRHEADER (default OFF) ON/OFF argument - toggels the format of R:
headers between a short
R:900920/1234 12345@KA2BQE.VT.USA.NA
and longer
R:900920/1234 12345@KA2BQE.VT.USA.NA [Underhill Ctr., VT] Z:05490
REPLYTO (default OFF) ON/OFF argument - toggles wheher or not to use
the REPLY TO argumenets for general messages orginated from the system.
Baiscally it works like this. If this flag is OFF, the variable called
reply_to is set to NULL if its ON its set to the sysop's home_bbs
field. Then if there is a home_bbs specced by the user creating a
message and it is not the same as this system, the it replaces the
reply_to value. When the internal header is created the "Reply-to:"
line is prited only if the reply_to variable is not NULL. This handles
problems of small personal BBSes with one or two other users that spec
a larger BBS as 'home' for ease of routing.
Changes to LIST
The List command and its brethren (K,E,R, etc) formerly would not
display any non-P message that had an @BBS filed. This has been
removed since we are doing almost all flood bulletins anyway. the only
thing that will not show is flood pointers('$'), unless you are the creator
of the orginal.
Changes to REQVER reply.
The format of the REQVER reply has been chnaged slightly with an aim
towards automated parsing of such messages in the future. THis is all
part of the reduction of R: and RFC-822 header materials to cut down on
overhead.
Changes to COMPRESS
The COMPRESS command now takes an argument. A string containing the
message types to be compressed T, B, P and special case $ and/or wildcard
*.
COMPRESS TPB$ removes all message types as well as flood pointers
COMPRESS * removes all types (functionally same as above)
COMPRESS B$ removes bulletins and any flood pointers
COMPRESS PB$ removes bulletins, personals, pointers (leaves any Ts)
Many of you may wish to hand clean the file by manually invoking
COMPRESS but may wish to put a COMPRESS $ in your event file to clean
out your flood pointers which can mount up on a busy system.
OUTSTANDING Problems:
None really - except I am still not 100% sure of the fix on the lockup
- simply because I am not sure what fixed it, only time will tell on
that. It has been running OK here for severa days and I can no longer
make it hang where before the circumstances were 100% repeatable.
There still is an annoyance of all those 'permission' denied messages
when doing things like K< , K@, K>, etc., you would not want to
initiate those actions as a remote sysop as it would clutter the air
with stuff.
I have not explicitly found the cause of the corruption of the arpa
format date that occurrs after long periods of system on the air. I have
seen it occurr in v1.45, but not yet in v1.46 - one workaround may be a
simple "M 0707 X 0" to cause the BBS to exit in the middle of the nite
and restart itself via looped PBBS.BAT This must be done carefully utilizing
the EVENT -S in CONFIG and EVENT files to keep the BBS from coming back in
and going to EVENT and EXITING several times before escaping the 'event
window'.
ROSEUSER edit sweeper does not do any H-address lookup on Home BBS, I
will be adding this shortly. I did clean up the extraneous redisplay of
user records in the edit mode though.
COMMENTS:
You will notice the new R: formats both do not have the 'z' for Zulu
time. Its is more or less a default. More and more systems have been
dropping the 'z'. It is not possible to run PRMBS on anything but Zulu
time, in reality. The TZ environmental variable allows you to run local
time and the tells PRMBS how to compute Zulu time to run itself. Zulu
time is assumed.
While LONGHEADER will let you run a longer header advertising your
exotic qth ('scenic sunny downtown Hoboken' ???!!##???) I would rather
see more select to keep LONGHEADER off.
The LONGIHEADER OFF will answer some complianst about 'extra materials' in
the message formats and at the same time will not compromise the features
of PRMBS whihc depend on it.
APPLICATIONS IDEA (1)
IF you have a station along way off that you wish to flood buletins
to, but not have them go all over beyond you area. Add a pointer to
your .FLD file and use distrib to encapsulate it in an RMAIL. For
example. Gordon, N2DSY wanted to pass NJNET stuff to me in Vermont from
NNJ. He takes his NJNET.FLD file and adds the routename NJNET to it.
Then he makes an NJNET.DST file that looks like
$RMAIL@KA2BQE.VT.USA.NA
NJNET
*** EOF
now you could add more RMAILs for other far flung systems as well, one
each and the '$' prefix so the BID gets embedded into the RMAIL and is
re-instated at the far end.
Then in your event file you have a command DISTRIB NJNET which will
take the bulletin pointer and embed it into a private RMAIL to go out.
The only problem may be if it goes through an AA4RE system along the
way which may have seen tha bull - but generally thats not the case,
since the whole idea is to be passing 'over' a zone where it may not
need to be seen. You may extrapolate from this that its best to have a
station on the neares border of NJNET's geographic border to the
desired target should be the one to do this.
APPLICATIONS IDEA (2)
To do some really heavy duty house cleaning you might do things like
KL -S SALE 123 or KL -S WANTED 123 , this says kill the last 123
messages that have SALE in TO/FROM/@/Title and ask me about it one by
one. You can always throw in the Rambo option ('-x') ad let it rip
without asking. "KL -XS SALE 123" (I pick 123 because most of us run a
few hundred messages and 123 is easy to hit and will get most of what
you want) You then could eyeball the killed by doing a LL -K 123 and
see if maybe you got a few you didn't want, then do an E -K #### an
salvage them if need be
PRMBS Ver 1.47 10/20/90
CONFIG changes - date/version
PROBLEMS fixed:
(1) If you used ""s in a '-s' search string it didn't always make proper
caseless checks and it failed to find message that it should have (fixed).
(2) When you did K@, K>, K<, KR commands messages that didn't match got a
message saying 'not permitted'for each non-match - annoying and wasteful on
the air (fixed).
(3) periodic contamination of the ARPA formatted date strings - which I
believe orginated in contamination of the table of 3 letter month abbrev-
iations. I altered the manner in which the LIST display lines are composed.
I believe it may have something to do with this. Since the problem was very
sporadic and could not be 'forced' to happen. I am only guessing here, but I
would bet more than a the price of a decent lunch that this should cure the
problem. (fixed ???)
(4) there is a periodic failure during logout by a user or BBS forwarding
system where the system bombs to DOS with an ERRORLEVEL 6 with a message
ulock: unable to open file ""
The variable for the userfile name is located close to the month table
mentioned above, they may be related. The strat of these occurrences is quite
close together. (fixed ???)
(5) Apparently a verrrryyy olllddddd problem. I have gotten random comments
about this for several months at least and never found anything ...
workaround had been to turn BBSMESSAGES parm off in TNC. BUT, there was a
bug and it was sitting right there for me to see, Ijust wass looking at it
wrong. If you were forwarding mail and you were waiting for the 'OK' and got
a 'connect request' - sayonara it just kind of sat there. Thanks to Bob
Nelson, KB1BD for bringing it up again and saying what ever magic incantation
it was that had me look at it yet one more time; this time with the right
perspective!
(6) The MID handling was faulty if a message came in and was address trans-
lated, the old MID wasn't kept and it would come in and dupe. The code was
corrected that if a message comes in and it has a MID and it gets translated
the old MID is now recorded as well as the new one.
(7) The code for reading in a '\' in front of a '#' to get sub-routes in in
the config and HROUTES file wasn't working as well as the H-Route code itself
was not working exactly as it should have. Its difficult to explain but it
now appears to be fixed. I have included a new sample HROUTES file based upon
NW Vermont, notice particulalry the top section with the local sub-regions
and specific BBSes connected to as weel as in the middle the section for New
York state, Northern NY sub-region, and the KD2AJ BBS in particular. I
believe these examples should completely show how H-Routes goes together.
#---------------------------------------------------------------------------
excerpt from CONFIG file showing my 'edit routes'
#---------------------------------------------------------------------------
\#NWVT.VT.USA.NA.HAMRADIO.ORG.ISO
VT.USA.NA.HAMRADIO.ORG.ISO
VT.USA.HAMRADIO.ORG.ISO
VT.USA.NA
VT.USA
VT
# soon these should not be needed, most pre-0.98 copies of PRMBS have
# finally been upgraded
USA.NA.HAMRADIO.ORG.ISO
USA.HAMRADIO.ORG.ISO
USA.NA
USA
NA
AMPR.ORG
*** EOF
#---------------------------------------------------------------------------
FILE: HROUTES.RS
# LOCAL Routing Information, based upon a NorthWestern Vermont (NWVT)locale.
# It has a few sub-region routes (SWVT,CVT,SEVT,NEVT) and explicit calls. These
# should be local only and they have no routing
\#SWVT KA1NZA
\#CVT N1BRT
\#SEVT N1BRT
\#NEVT N1BRT
NTSVT N1BRT
WA2SPL WA2SPL
N1GMU N1GMU
W1KOO W1KOO
KA2UZC KA2UZC
WB2MIC WB2MIC
KA1NZA WB2MIC
# Continents in general
OC DEFLT
SA DEFLT
EU DEFLT
AU DEFLT
AF DEFLT
AS DEFLT
AN DEFLT
#USA - States and territories
AK DEFLT
AL DEFLT
AR DEFLT
AZ DEFLT
CA DEFLT
CO DEFLT
CT NEBBS
DE MIDLAN
FL DEFLT
GA DEFLT
HI DEFLT
IA DEFLT
ID DEFLT
IL DEFLT
IN DEFLT
KS DEFLT
KY DEFLT
LA DEFLT
MA NEBBS
MB DEFLT
MD MIDLAN
ME NEBBS
MI DEFLT
MN DEFLT
MO DEFLT
MS DEFLT
MT DEFLT
NC DEFLT
ND DEFLT
NE DEFLT
NH NEBBS
NJ MIDLAN
NM DEFLT
NV DEFLT
# Note here that the Hierarchy is TOP DOWN - NY the NNY then BBS in NNY
# Basically I connect direct across Lake Champlain to KD2AJ so I name him
# explicitly. The NNY region goes to my default route to the main Mail BBS here
# in NWVT, and NY in general, outside of NNY is MIDdle AtLANtic routing
# I placed in the KD2AJ.NY in just in case mail came in without the NNY.
#look alittlelower and see a similar example with two nearby Canadian Stations
NY MIDLAN
KD2AJ.NY KD2AJ
\#NNY.NY DEFLT
KD2AJ.#NNY.NY KD2AJ
OH DEFLT
OK DEFLT
OR DEFLT
PA MIDLAN
PR DEFLT
RI NEBBS
SC DEFLT
SD DEFLT
TN DEFLT
TX DEFLT
UT DEFLT
VA DEFLT
VI DEFLT
#VT - dats my state!
WA DEFLT
WI DEFLT
WV DEFLT
WY DEFLT
# Canada - Provinces and territories
CAN DEFLT
AB.CAN DEFLT
BC.CAN DEFLT
MB.CAN DEFLT
NB.CAN QUEBEC
NS.CAN QUEBEC
ON.CAN QUEBEC
PE.CAN DEFLT
PQ.CAN QUEBEC
VE2FKB.PQ.CAN VE2FKB
VE2CSC.PQ.CAN VE2CSC
SK.CAN DEFLT
-----------
ALTERATIONs
(1) Due to the introduction of 'expired' as a LIST type option (see below
for more detail), the infamous 'RAMBO option, which basically works only in
message KILL, has now become '-Z'!
(2) The list, read, edit, kill commands using @, >, <, and R had an
awkward syntax. It has been cleaned up. For example tolist all messages from
KA2BQE dealing with RATS in the range from 12000 to 14300, you would type;
OLD SYNTAX: L< -s RATS KA2BQE 12000-14300
NEW SYNTAX L< KA2BQE -s RATS 12000-14300
(3) The KX and EX (expired) and KP and EP (already read private)
commands have been merged into the normal list commands they become '-x ###'
and '-p'. This is more flexible and allows more selectivity. The'x' command
is not limited to bulletins, but may be used for all types using the 'bhkpt'
options to refine it.
To list all bulletins 30 days or older from N2DSY and on the subject of RATS
type:
+---+----------- expired 30 days
| |
L< KA2BQE -xbs 30 RATS 1
|| | |
|| | +--- start at message #1
|+----+-------- search for string "RATS" in title
+-------------- bulletins only
The parsing of the line is now more sophisticated and flexible.
(4) There are three additional SETPARM commands. One called MAXLIST. It
defaults to 0 value - when set to non-zero it limits the number of messages
that can be listed in one command execution by any on the air user. If a
REMOTE Sysop absolutely needs to have a longer display, he can flip to
REM_SYS mode and alter the MAXLIST parameter.
The second is FORWARDKILL whihc defaults to OFF. When set ON, it will the
text file of any successfully forwarded message immediately after forward of
that message. This cannot be used if you are doing any archiving. The
MAIL.DAT entry for the message is still there and must be removed via
COMPRESS.
The third is FLOODKILL, if the FORWARDKILL is set ON this flag has no
effect, but if FORWARDKILL is set OFF this flag may be set on to remove '$'
flood pointer text files (they are 20 bytes or so long but use a 4K
allocation) immediately upon being forwarded and killed.
(5) When calls are added to the 'L' port list showing who has connected.
The comparison willnowbe made including the PORT the user ame inon. Before
if a user came in on port B and then on port A you saw only one entry for
port A. Now you will see the most recent connection for a given call/ssid
for each port
(6) In recent prior versions the 'L" command was set up so a new user
typing just 'L' would get re-mapped to a "LL 20" to rpevent several hundred
messages from being listed. This mapping is no longer done becauseofthe
SETPARM MAXLIST (see #4 above).
(7) After a discussion with a number of users and sysops, the '-s' search
option in the lstmsg() commands is now aimed purely at the title. It no
longer examine the TO FROM AT fields. Since the string can be used with L>,
L<, and L@ It provies more flexibility and is less likely to 'find' unwanted
messages to search title only. I belive the number of you would want TO FROM
AT TITLE searched simultaneously for a specific string is far less than
wanting to look through title only and TO FROM AT individually.
(8) Some slight changes have been made to the handling of the smart system
Id ( the [SID] ) A line containing a SID may be accepted at anytime. Once a
SID line has been accepted, subsequent SIDs may also be tendered, even the
'dumbell' SID [XXX-0] ('0' implying system is 'smart enu to know its dumb,
ONLY!), but there is no going back to being a user. Once any SID is tendered
ANY ERROR of any kind will cause the system to be IMMEDIATELY DISCONNECTED.
The theory is that the SID is an indicator of talking to a computer and
'computers never make mistakes' (yeah I know!!!!!). All of this is being done
as an attempt to keep mail from being lost by a number of strange happenings
I have seen when different BBS systems get 'out of sync' with each other. It
is also a part of some new agreements on standards between the various
codewriters.
A further refinement of this is that while PRMBS accepts 'extended
reponse', menaing OK/NO/LATER/REJECT, it will check only for the letters O N
L R and will consider any other response from a system as an error and
generate a disconnect.
(9) For you BPQ fans out there. I have added yet another feature which was
sorely needed. As we all know once a message gets out inot the TNC often the
propagation delays from NETROM ect make you have to use inreasonably long
timeouts to keep from dumping a connection thats almost there, but the
chnannel is a bit busy. If you designate a port as being a 'BPQ' port (Q port
descriptor) now not only does a 'B' command by a user put him back into the
node, but for any transmissions, that is long lists by user, as well as long
messages being forwarded. The timeout timer will keep being reset until there
are no unACKed packets then it starts counting down! There is an extended
COMBIOS call in BPQ whihc lets me query for unACKed packet count.
PRMBS Ver 1.48 01/04/91
*** update version for anyone using v1.45, 1.46, 1.47, see included RME
files for these versions as well ***
CONFIG changes - date/version
PROBLEMS fixed:
- in the simple 'K' command (no modifiers ><@) the '-x' command did not
work correctly - fixed.
CHANGES:
Flood message - pointer now has same message number as the master bulletin
and no text file is generated. These '$' flood pointer messages will now
appear numerically out of sequence to the sysop doing a LIST command. Any
command taking a message number to search for will not 'find' a '$' if it is
specified as a staring number (i.e. if you say 'K 1234-1250' and 1234 is a
'$' it will get the next message; on the other hand you may simply say E
1234-1250 and 1233 will be found and you 'step thru' to 1234 and kill it)
MAILWORK has been modified to handle out of sequnce MAIL.DAT records, as has
bldfwd().
In general message flooding is now several orders of magnitude quicker, and
will be noticeable to XT users. Older egenrated bulletins will forward out
OK.
Speaking of bldfwd(), since its only purpose is to produce 'Mail For:'
strings, I added a 'KB abort' call so that if you are bringing up the board
and want to get at it quickly to do console work, you can hit ESC and it will
bypass that long run through the MAIL.DAT file.
COUNT command
The count command is simply the list command without the detail line display.
It allows you to say count or C> , C<, C@, cr, or cp, with any modifiers that
are available to LIST and display the count of messages that satifies the
criteria you used. It is available to all users. the HELP.RS has been
modified for COUNT. You may simply excise the paragraph on Count and insert
it into your file - no other changes were made to help except to the hELP ALL
to add count to the list of available helps.
OUTSTANDING PROBLEMS:
ulockf: unable to open ""
This is still there, not so often now - will keep looking for it. I have
had it happen once, as had WA2NDV in 10 weeks or so ... not so often but
still with us.
DDOS/DV - multiple copies of PRMBS can generate dupe bulls --
Problem is for you - it will not propagate - you may recieve the same
bulletin simultaneously in every window, so as it hasn't saved, the BID will
not be rejected. But once you try to forward out all the dupe pointers they
will get rejected, all but the first - so its an annoyance not a serious bug.
It happened here several times in 7 days then not again and had never happened
before. the fix is non trivial - I will not implement it in the single user
copy - it will be done in the multi-connect copy currently under design.