
Friday 19-Nov-93 17:00:09

--1300--

I want to have "the reading of gfiles from the gfile-menu" logged.

CONFIG's READTEXT and READFILE don't seem to do it.

Is there another flag?  Or should READTEXT be covering this?
(It seems to cover news:files, but not gfiles.)

How do I tell who (if anyone) is reading the gfiles?

Hey, I think I found it.
Create a NEW CONFIG/LOG called ReadGfile with a log-id # of 408.
When you read a gfile, it now gets logged.

(I don't know if I accidently removed that CONFIG/LOG entry, or Ken
did.)

Now, if I could just get it to log NEWS-files with a "ReadNfile"
marker instead of with the general-purpose "ReadText" entry....

> ReadGfile
> ReadNfile (not supported?)
> RunPfile
> ReadText  (all text-files read that do NOT fall into the above 3 groups.)

I took several other missing log-id #s and filled them with foo1 foo2 foo3
strings.  They too are missing from config/logs, but are appearing in the
logs.  (Now I just need to find out what they do.)

118 NetMail
224 ???
313 FullMaint
408 ReadGfile
409 ???

Any help?

Hey, I found #118.  It's logged whenever netmail is sent (MS name@z:n/f.p)

Any luck on the others?

Also, is there a way to get CNet to log that an event happened?
Something like:
> 18-Dec 21:34 Port 0    RunCNetC pfiles:BBS/amaint
> 18-Dec 21:34 Port 0    RunCNetC pfiles:maintenance/pointers

(Indepentant of whether or not the activity that occurs also writes
its activity to the log.)

--1301--

When using VOTE's "Xamine" choice, REAL-NAMES would be much more
informative here than a long list acct #s.

Showing which callers voted FOR or AGAINST which topics, instead of
having to look-up the names for acct #478, #582, #611, etc...

--1302--

"Browse Global MESS BY #" as well as the current "Scan Global MESS BY #"
 ^^^^^^                                           ^^^^

Browse, and pick which mess's you wish to read, as you go along.

--1303--

When you type HELP;SYSTEM the last 4 lines display are:

> If you are using an ANSI terminal, you may use the cursor up and down
> movement keys to review the last 10 commands that you've entered.
>
> Enter another topic (or "MENU"):

But from the HELP-prompt you're at, handy "cursor up/down history"
isn't available.  History-cursor would be just as handy in HELP as
it is elsewhere.

--1304--

I also notice there isn't a sysText:help file for UDBASE or FILES or
FILEAREA or any kind of file-area overview, like there is for HELP;BASE.

--1305--

After marking many files for D/L, I frequently find myself executing
"SS" and "CR", "SS" and "CR", "SS" and "CR", "SS" and "CR" as I
check my CREDITS and compare that with how many files/bytes I've
marked/unmarked.

How about adding that info right into "CR".

OLD>                     Files        Bytes
OLD> --------------- --------- ------------
OLD> Your uploads  :        96       2860 K
OLD> Your downloads:       217      19998 K
OLD> Default ratios:       4:1         20:1
OLD> Your credits  :        93       3904 K
OLD> Usable today  :        25        900 K
NEW> Selected      :        12        657 K

Or better yet, squeezed into the last line of SS/DS:
OLD> 6 Total         1232K   12/  657K    Time: 9:30  Cost: $1.234
NEW> 6 Total         1232K   12/  657K    Time: 9:30  Cost: $1.234  Max: 25/900K

Or both places.

--1306--

When a user leaves "FeedBack" (102), it is logged as "SendMail" (107) instead.
I have both of these set to the same under CONFIG/LOGS.
How do I get feedback logged as "FeedBack".
And mail sent as non-FeedBack as "SendMail".

--1307--

I don't care for BROWSE to "always list x items".
Why not obey the screen-height that I specifically set in "ET"?

--1308--

I always wanted to get a hold of the classic source code for "ELIZA".
A pseudo-artificial-intelligence game.
Ask questions, make comments, and the program replies as best it can.

I wonder how long a user would think he's chatting with another user before he
realized it's just a pfile?  Or replace "sysop-chat" when you aren't home,
with ELIZA.

ELIZA adds your coments to her database, getting "smarter" and "smarter" with
each use.  I wonder what she would do after 5000 callers had chatted with her?

Anyone have the C source to ELIZA?

--1309--

I received a piece of netmail with CNet v2.94 and CNetToss v2.96a.
Under "MR" the header looks like this:

MR> Item: 27 (of 27)
MR> subj: CNET Echo
MR> From: Mario Bonelli  (Mario Bonelli@1:273/934.0)
MR> On  : Sun 14-Nov-1993  8:32p



But when I YANKed it into a file, the header looks like this:

YANK> mess: 27
YANK> Subj: CNET Echo
YANK> From: Mario Bonelli
YANK> To  : yyy  Dx (junk)
YANK> On  : Sun 14-Nov-1993  8:???

Junk chars in the "To:" field.

--1310--

When killing pfiles/gfiles/news-files the 2 prompts are VERY misleading.
> Delete file(s) also [Yes]?
> Kill dh0:filename are you sure [Yes]?

They should be:
1> Delete file "pfiles:bday" [Yes]?
2> Delete item "Show who's birthday is today" [Yes]?

Show the file-name when asked about the file.
Show the item-name when asked about the item.
(NOT the current method, "don't show the filename when asked about deleting it"
then "show the file-name when asked about the item".)

Also, asking #2, then #1 is more CNet-like, similar to other KILL prompts.

--1311--

Under CNet v2.94...

Perhaps someone will log-in under a NON-SysOp acct and verify this
before Ken goes looking for a bug...

I use "EL" to set-up a sub as:
> Private messages    : Force

Yet users can still answer "[No one]" at the prompt marked:
> Enter the ADDRESSEE'S handle [No one].

(If you are FORCING private msgs, why would you want them addressed
to "[No one]"?)

In fact, anytime (forced or non-forced) the user answers YES to:
> Private message  [Yes]? Yes
then
> Enter the ADDRESSEE'S handle [No one].
should DEMAND a handle, forbidding "[No one]".

Maybe like:
> Enter the ADDRESSEE'S handle [mandatory].

--1312--

Some users started talking about their dislike of the way CNet re-
lists files that have already been listed during BROWSE/marking.  (I
assume this is to re-display the filelist as marking can cause
it to scroll of the top of the screen.)

Maybe have the re-listing work exactly as it does now, but with 1
exception...
Those files that are re-listed have the file # shown in a different
color.

Or if you want to get rid of the actual need to re-list at all, maybe
add something like:
> BROWSE: [D]ownload, [*]mark, [E]xamine, [G]rab, [Q]uit, [R]ead>  *4-10!
The "!" will suppress all output and just do a carriage-return
(not a line-feed) and the screen won't scroll AT ALL.

--1313--

M> Anyone have a Good trapdoor.cfg for a 14.4.  I seem to be getting a few

If you dial the BBS at 313-473-2020 you can log-in with the name "DEMO" and
password "DEMO" without creating a new acct.

You can then view/download the actual trapdoor.cfg file in use there.

It's online under the gfile-menu marked "config-files in use here".  (Many
others too.)

Sure wish 100 other SysOps did this, too.

--1314--

Ken should allow comment-lines in BBSTEXT.

M> I think that if He did list what line did what, It would change so fast
M> that the current 130+K Readme file would pale to it.

That's the beauty of comment-lines right in BBSTEXT.
As the lines move, change, add, remove, so do the comments.
Your comments, or Ken's, or I'm sure someone will U/L their's too.

It would be handy even if Ken added 0 comments.
Just so that comment-lines would be possible.

They, of course, would be stripped off as BBSTEXT was read into memory, so
there should 0 bytes of extra memory used.

--1315--

> Checking for duplicate U/Ls is too slow on 50,000 file systems.

How about if CNet did the dup-check as a spawned background task?
The user could start U/Ling immediately, but the U/L would be aborted/deleted
if it was then determined to be a dup-file part way through.

--1316--

> Locating random mail-send-to-null bug:

Some things to watch for.  Is this happening when:
1) the user's mbox full?
2) the sender is hitting (or near) his max msg/call limit?
3) sending to handles VS real-names VS acct #s?
4) the receiver is online at the time?
5) the user's mbox is empty?
6) the sender is hitting (or near) his max editor-lines limit?

> The first two DID NOT
> SAVE into his mailbox (!!!!) although the system SAID that they did.

Also, is the "mail-sent" string located in a part of the code that verifies
that the send did, in fact, go OK?  Successful open, write or append, and
close?

--1317--

I'm running an Arexx script from the CLI.  (Not from online.)

0> address CNETREXX3
1> 'dropCarrier'
2> 'dropCarrier'; 'resetModem'
3> 'addTime' "-1440"
4> 'bbsCommand' "O!"
5> 'addKeys' "O!"

I wish to drop-carrier in a normal fashion, as if the caller just
logged off normally.

All of the above 4 methods have (1 or more) side-effects that I wish
to avoid:

> Logging "SIGNOFF: LOST CARRIER" or "TIME LIMIT EXCEEDED".
> Displaying to user "Time limit exceeded".
> Doesn't work from ALL prompts.
> Only takes effect AFTER the arexx-script exits.

Is there a better way to do this simple task?

--1318--

DW> Are you saying the "Repair_Sub" file in 2.63 can also be used to weed
DW> posts and responses(good ones) from subs as well?

MA> It doesn't weed the responses.  What it does (as near as I can tell), in
MA> addition to the above, is remove a *mess* of garbage that Ifido puts in to
MA> your main TEXT files that Amaint doesn't remove for some reason.

I have NOT seen "Repair_Sub" described by Ken/Jim/anyone as use for that
purpose.  (And what is a "main TEXT file"?)

It does shrink your sub's _Text files for it is actually removing the
previously amaint-killed msgs.

--1319--

Is anyone familiar with a prg available on the IBM called AIM?

I'm not sure how it works, but gather it is similar to automatic
AreaFixing.  Something like:

You post a msg in a certain Fidonet msg base set aside for the
purpose, with the tagname AIM_SEARCH (or whatever).  The msgs would be
something like:

> From: Bill Beogelein@1:2410/207
> To  : AIM         (Or "AIM 1:123/*" if you just wanted to search some BBSes)
> Subj: 1:2410/207  (So that all msgs from this system are grouped together)

And the body of the msg would be something like:
> FILE ar13*            (To do wildcard searches based on filenames)
or
> DESC amiga*reports    (To do wildcard searches based on descriptions)

Each BBS that received the msg would automatically search for the
files and post a reply to that user like:

> Your AIM file-search on 25-Nov-93 for "ar13* found the following 4
> matches at 617-555-1212, 1:123/456, Boston, MA...
>
> * Text Files/Amiga Magazines
>
>   # U/L'ed DLs Filename   EXT Size Description
>  == ====== === ============== ==== ===========
>   1 23-Nov  16 ar134      LHA  81K v1.34 of AmigaReport 19Nov93
>   8 06-Nov  13 ar132      LHA  66K v1.32 of AmigaReport 05Nov93
>   9 01-Nov  10 ar131      LHA  73K v1.31 of AmigaReport 29Oct93
>  12 25-Oct  21 ar130      LHA  61K v1.30 of AmigaReport 22Oct93
>
> [BBS system info, FileRequ'ing procedure, # lines, bps-speed, etc.]
>

If you didn't want your system to search/reply, (or if your non-CNet-
BBS didn't support this auto-searching), you could still read/write to
this echo, but with your system's search/reply AIM-flag turned off.

You would basically be able to search for any file, by any name or any
description, on any BBS in the world, by posting one (1!) msg on one
(1!) BBS down the street.

--1320--

S> will ONLY accept a port# not a handle.  I think it would be a good idea to
S> allow the same information (handle/uid, port number) at all the applicable

You forgot "real-name".

The whole ball of wax would be:
*EVERY* CNet prompt that accepts any 1 of the following, will also accept the
other 3...
1) real-name
2) handle
3) acct #
4) port #

*EVERY* prompt that accepts any/all 4 will do so with the same set of rules.
(And performed in this order.)
A) The "#" char must proceed all acct #s.
B) The "p" or "P" char must proceed all port #s.
C) Handles
D) Real-names (if the user doesn't have this marked as PRIVATE.)

All online prompts, and all offline CNet cmds, everything, would work like that.

(I know the "#" and "P" ideas aren't going to be popular, but I can't imagine
how else they can both be used.  Maybe "#" could be optional, and presumed to
be an acct # if the "#" char isn't present.)

--1321--

S> Hmmm the only way I can think of to safely have comments in bbstext
S> without screwing things up is a MCI command.

Yes. I suggested the MCI code "\;" be used for comments, ages ago.
1) ";" is still available.
2) ";" is a commonly used comment-char.

S> But then Ken would have
S> to write a routine to strip these MCI comments from the BBSTEXT as it is
S> loaded to keep them from taking tons of ram.

Should take 1-2 lines of code and 1-2 mins to write.  (I'll gladly do it.)

If it was done as an MCI code, it would be available everywhere and
Ken wouldn't have to keep adding comment-support on a file-by-file
basis.

--1322--

SubAdd...

Bob, this is definitely an "include-with-CNet-on-the-3rd-extras-disk"
type utility.  (Now if we can only convince Ken to START one.)

Please consider changing things like:
> Could not open your newsgroups file
INTO
> Could not open your newsgroups file:  "UULib:Newsgroups"
We INSTANTLY know the full-path&name of the offending file.

Also, please consider including a small, sample UULib:Newsgroups file
with this archive.

Also,
1) Bob uses version strings like "SubAdd 1.0 (11.28.93)"
2) CBM suggests version strings like "SubAdd 1.0 (28.11.93)"
3) I suggest version strings like "SubAdd v1.0 (28-Nov-93)"

#2 is confusing, and #1 makes it worse.
(Only #2 is "officially sanctioned" by CBM.)

If you use SAS's __DATE__ macro, you don't even have to worry about
forgetting to update the version-string date.  The compiler does it
for you.

> #define VERSION "$VER: SubAdd v1.0 " __DATE__ ", "  __TIME__
Gets you "SubAdd v1.0 Nov 28 1993, 22:33:13"

Also, any plans on doing a FIDONET version of SubAdd, too?
(100s of Fidonet echos aren't fun to add 1-by-1 by hand, either.)


BA> Also, please consider including a small, sample UULib:Newsgroups file
BA> with this archive.

TB> Hadn't thought of that, since anyone using this utility likely already has
TB> one.

I'm not running USENET yet, but would love to, soon.
But I'd like to have everything 100% set-up BEFORE I throw the switch.

BA> Also, any plans on doing a FIDONET version of SubAdd, too?
BA> (100s of Fidonet echos aren't fun to add 1-by-1 by hand, either.)

TB> I don't run a FIDO system.. I have _no_ idea what kind of FIDO equvilants
TB> there are to the UUCP file setup (IE is there some master file that tells
TB> it what groups you get?  What format is it in?  Etc)

I have NO idea if it's official, or not, but I always get my Fidonet list
from my HUB in a format like:
> AMIGA            Amiga International Echo
> AMIGAGAMES       Amiga Gaming
> AMIGASALE        Amiga NON-Commercial Items for Sale
> AMIGA_CDROM      AMIGA CDROM & CDTV Discussion/Sales
> AMIGA_COMMS      Amiga Communications Software and Hardware
> AMIGA_LC         Amiga Language C programming & SAS/C Conference
> AMIGA_MUSIC      Amiga Music/sound topics
> AMIGA_NET_DEV    Amiga Network Developers.
> AMIGA_OS&EM      Amiga Operating Systems and Emulators

So, I can only assume you could use exactly the same code as SubAdd,
but just read "TAGNAME" and "DESCRIPTION" from a "CNet:Fido/FidoList"
(or whatever) file.  (Exactly why I wanted to see what a UULib:Newsgroups
file looks like.)

And, of course, an initial prompt like:
SubAdd> Create [U]seNet or [F]idoNet subs:

BA> AMIGA            Amiga International Echo
BA> AMIGAGAMES       Amiga Gaming

TB> In the above example, how would 'AMIGA' and 'Amiga International Echo' be
TB> translated into CNet?  (I don't know what the data-file pointers are in a
TB> FIDO subboard...)

I'm guessing at all this CNet programming stuff myself.

Put 'AMIGA' into a sub's "Name used on disk (unique)" field.
I'm not sure which it is without experimenting...
} NewSubboardType.SubDirName
} NewSubboardType.DataPath
} NewSubboardType.ZeroPath
(If it was called NewSubboardType.Unique (just like the online-prompt), I
could tell immediately which one it is!!!)

Put 'Amiga International Echo' into a sub's "Title to appear on List" field.
} NewSubboardType.Title

SubAdd would also need prompts like:
> What is the name of the Echo-List file: SysText:EchoList.txt
> Put these areas into which subdir #   : 231
(Where users could specify a physical sub #.)

--1323--

JS> Well, we had the Cirle MUD up here on Future World all day yesterday,
JS> and I must say that it is IMPRESSIVE!  At one time we have 3 users online

Jim, are there any other pfiles current running on FW?
Does everyone have access?  (I don't)
Are there time-locks in effect?
If so, what times?

(It would be great if CNet showed these time-lock times instead of just
saying, "you can't enter, and I'm not tell you when you can".

OLD> You may NOT enter that subboard at this time.
NEW> You may only enter this subboard during the hours of 4-7am.
(CNet obviously knows, it just doesn't tell the user.)

--1324--

Please "EL 1-99" and toggle that flag allowing CA.
(You'll have to do it within each sub-dir until Ken adds a
new option to EL that allows things like "EL 1-99 ALL" meaning
all subdirs, too.)                                ^^^

--1325--

D> I got them inporting just fine, but
D> the LONG description just shows up as another file.

I hate that.  It doubles the # of files online.  (I have too many already.)

I'd love to know how to add items and enter short OR long descriptions into
CNet from the CLI.  I'd GLADLY write a tool to do this.

(And NOT using "AO" or waiting until the next amaint run.  It would import
tick-files immediately.)

Currently I don't think it is possible.  I see addItem() and addDescription()
and AddFileNote() calls in cnetfuncs.h, but no docs on how (or even "if") we
can use them ourselves.



Also, what version of tick are you using?  The latest one online here is a 1.5
YEAR old beta release.  Please U/L the latest.

> Upload Rsp File           Size Description
> Jul'92   0 ftick090   LHA  22K FTick v.090 For Tic Files!

--1326--

When users decide to leave feedback from the log-in prompt, what are
these prompts for:
1> Return a receipt if no reply [No]?
2> Return original message also [No]?

Can they be removed?
You can't really do #1 or #2 for feedback left during log-in, can you?

--1327--

Uploaded 03-Dec-93...

UL> Lists names and latest version numbers of 67 programs written by Bill
UL> Allen Beogelein as of 01Dec93
UL>
UL> Do you have the latest versions?

Do other pfile authors have a list of filenames and version numbers of their
work?  If so, please U/L.

Once per week, I automatically run a DOS-event like:

> Delete Gfiles:MyList
> List >t:tmp Pfiles:Mine/ LFORMAT "c:Version >>Gfiles:MyList %s%s FULL FILE"
> Execute t:tmp

to automatically create my versions-list.

--1328--

DS> What do I need to do to set this right... Do I need to set something so if
DS> a users balance is NEGATIVE that they will still be allowed to logon!!!!!

> AC
> Detailed [Yes]? No
> Connection time charges       $  0.0075
> Other activity charges        $  0.1000
> Total cost of this call       $  0.1075
> Remaining acct balance        $ 26.5525
> Minimum balance allowed       $  0.0000

CONFIG/ACCOUNTING allows you to set "Minimum balance allowed" to $x or -$x.
(per each level, and per each schedule that you are running.)
I don't know if that's what you are looking for, or not.

Also, In all the BBSes that I've called, I've seen VERY few using the
accounting features at all.  Maybe a few bugs still live in there.  (I
found a few, but they've since been fixed.)

Maybe we should all (including FW) attach a few charges onto various things.
(But NOT have them effect a user's ability to log-in, or drop his access
level, etc.)  Even if you run a 100% non-charge BBS.  (Just keep a dummy-total
of "this is how much you COULD have been charged, for these services, if this
had been a pay-BBS".)

I don't know how bugs can be found/fixed in a section of the BBS that
we never use.

--1329--

P> A release date.. You gotta be joking... As far as I can tell we are
P> no closer to having a bug free version than we were 3 months ago.
P> Its been so long since we've seen a bug fix release that I cant
P> remember how long its been.

Full update v2.96 was released 21-Oct-1993
Partial update v2.96a was releasd 23-Oct-93

I still feel some kind of release date estimate should ALWAYS be available to
those that have been incredibly patient for so long.

Just mentioning "Jan 94" or "Jan-Feb 94" or would be better than not knowing
at all.  I know things can be VERY hard to predict, but Ken/Jim should be able
to hit at least a 2-3 month estimate-range.  If you have know idea if v3.0 is
going to be released tomorrow, or by 01-Mar-94, that's pretty bad.
I think MANY users ARE thinking, "maybe it'll be released tomorrow, it's been
delayed since 'end of summer 93'".  It just further disappoints them when it
isn't.

If you are awaiting the arrival of a flight containing a loved one.  Which
would you rather be told about their flight?  It's going to be:
1) 5 mins late. Then 5 more. Then 10 more. Then 5 more. Then 15 more.
   (So you sit with your nose pressed up against the window.)
2) Or just, give me the truth, from the start.  It's going to be 40 mins late.
3) Or, "no comment".  Just wait forever.
4) The plane has crashed.  114 killed.  91 injuried.

I'm a big boy, I can handle the bad news.  But just tell me SOMETHING!

CNet is currently running at #3.  The flight is MONTHS behind schedule.  And
people are beginning to wonder if #4 has happened.

--1330--

WC> ...and not only that, but SAVING OLM's to your mailbox doesn't seem to
WC> work on my system.. Dunno why... any ideas? Anyone?

Is the mail-box full, forwarded, locked, etc?
Does the sender have a send-mail-flag ON?
Does the receiver have a receive-mail-flag ON?

--1331--

> Encrypting password files...

All good, 1993, state-of-the-art BBS software should have slick
features like that.

I'll even give you the file that contains my password to access my
online DOS-shell.  You won't be able to do anything with it.  It's
encrypted with an additional 2nd password.

Oh yeah, I'll give you that 2nd password file, also.  You *STILL* won't
be able to do anything with it.

Oh yeah, even if you have the executable, and even the C-source code,
for both the encrypt/decrypt methods.  You **STILL** won't be able to
crack it.

I know it sounds too good to be true, but it can (and is) possible.
1) I give you the encrypted file.
2) I give you the file that contains the password I used to encrypt #1.
3) I give you the 2nd file that contains the password I used to encrypt #1.
4) I give you the executable I used to encrypt #1.
5) I give you the executable I can use to decrypt #1.
6) I give you the C-source code I used to encrypt #1.
7) I give you the C-source code I can use to decrypt #1.

Even with *ALL* 7 items, you won't been able to decrypt #1, or even
tell me what the passwords are in #2 and #3.  (They too are encrypted
with the same method!!!)

The C-source that I use, is just 5-10 lines of short, simple, fast code.
The small, executable takes less than 1 second to run.

It's a great, clever, setup!
I guess some BBS programs are just better/smarter/safer than others.
No names, please.

--1332--

FW NEWS> Change all your passwords...
FW NEWS> Change all your passwords, again...

Jim, has this hole be 100% plugged?

Even if it has, Ken should use the exact same remote-shell-password
encrypting techniques to encrypt the user's-password-file, too.

Even if someone D/Led SysData:bbs.udata and got everyone's password,
they couldn't do a thing with it.

Maybe some idea could be obtained from:
The Cryptography BBS, (703) 237-4322 McLean VA.

(If it's still open.)

If anyone calls, please get any/all types of password encryptions C-source
code examples.

I'm not sure why it's "not exactly".
That pretty DOES seem what is occurring here.
You can't figure out the RIGHT passwords, but your substitute WRONG
passwords will work also.  No?  "Not exactly"?

Also,

I guess anytime 1-BYTE + 1-BYTE = 1-BYTE methods are used, there won't
be that many different guesses you'd have to try.

What if the encryption produced something like:
Each 1-BYTE + 1-BYTE = 16-BYTEs of data output?

Then a 4-byte password + 4-byte keyword would give a 64-byte result.
Your "chart" would have billions of more combos to try and guess.

Also,

What if CNet kept it's current method, but put a stop to the
"wrong passwords might also work" loophole?  Maybe use a 32-64 byte
CRC checksum method, so that the RIGHT passwords and the guessed
passwords word have to be exactly the same 99.9999999999% of the time.
(Even if they both produced exact duplicate sysdata:passwords files.)

Also, what if CNet didn't "chop" keywords and passwords down to equal
lengths, (shortest winning out), but padded them to equal lengths?  A
password of WILLIAM and a keyword of BOB would become WILLIAM and
BOBOBB.  Or maybe kept padding them until both reached there maximum
lengths.  WILLIAMWILLIA BOBBOBBOBBOBB

What if CNet filled all the unused 0s in sysdata:passwords with randoms
characters?  (Just to make it harder to tell where words start/ended.)

What if CNet tossed in encrypted parts of my CNet serial number?  Or
my whole CNet registration key-file?  (So that even using the exact
same passwords would produce different sysdata:passwords files on
different systems?)  Or you'd need a copy of my registration key-file.
(If you already have that, you've already broke into my BBS.)

What if *ALL* of the above methods were used together???

I guess what I'm really asking is:
What would have made your cracking-job 10000x times harder/longer than it was?

--1333--

> Ctrl-X
> Abort Check-spelling Get Help LineEd New Put Quote Save Zip-UL?

How do I add/remove/change the cmds available from the full-screen editor?
I want to add a pfile.

I can't find a BBSMENU entry.
(Don't tell me there isn't one.)

--1334--

> I assume you directed that post to "Sysop" so that every
> sysop would think the message directed at him or her.  Even with those
> familiar with MCI, that kind of message can make one's heart skip a beat

Yes, I agree.

I've been trying to get Ken to make "the ability to post network msgs
addressed to 'SysOp'" into a sysop-only function.  (Very useful, but
its power should be limited/controlled.)

But I don't think he'll do it until the networks are flooded with such
msgs, every sysop thinking that they are all addressed personally to
them.

--1335--

BB> So that my arexx-pfiles can be run by users online with cmds like:
BB> > MyCmd Bill
BB> > MyCmd Bill Beogelein
BB> > MyCmd47
BB> > MyCmd 47
BB> and my pfile will receive the "Bill", or "Bill Beogelein", or "47"
BB> strings in all four cases.

NO> This will NOT work in call 4 of your cases, but it might help somewhat...
NO> getuser 1307358;params=strip(result)

Exactly what field of cnet.h is this looking at?

NO> The cases it wont work on would be using a long variable - I forgot the
NO> maximum length, its around 15 or so, and the command is included in that,
NO> everything else is trunctuated, also I dont think MyCmd47 would work with
NO> this method, though I'm not possitive.

I really need it to work in all 4 cases.  And all lengths.

NO> If there is another way to do this more
NO> correctly, I would definitely LOVE to hear about it!

Anybody played with these:
> char  pitem[8][61];     /* last input parsed list */
> char  GBuffer[16][81];  /* General purpose 2-D text buffer */
> BackBuff [10][80],      /* last 10 commands executed */
> Pathx     [8][64];      /* g/p/n subdir history */
> CStack       [80],      /* Command stacking */
> KBuff       [256],      /* keyboard buffer! */
> ABuffer     [256],      /* workspace */
> InBuffer    [256],      /* after a line input */
> UBYTE hmmmm[3];         [no comment]

I guess the "really, really correct" way would be to get Ken to see the
importance of this and add a new fixed, "GetUser 99" variable that would never
change as CNet is updated over the years.  (A 'movement' in the structures
(effecting 1307358) could really cause a powerful arexx-pfile to go wacko on
your HD.)

If anyone wants their arexx-pfiles to work online like real CNet cmds
do, just call this arexx subroutine.  The variable "args" will then
contain whatever the user has enter from online.

GetArgs:
   args=""
   do i=1302246 to 1302246+(61*8) by 61
      'getUser' i; j=RESULT
      args= strip(args j)
   end
return args

--1336--

R> Yes, you just have to do it properly.
R> BBSCOMMAND "bbs 1 -f -B%b -b%B"
R> That should work for you.

Works here.

It is unfortunate that Trapdoor and CNet use exactly the opposite options
representing "b" and "B".

--1337--

On 3 different occasions, that number displayed in the CONFIG/ARCHIVER
window "(# 4)" became junk-chars during heavy use picking/editing
CONFIG/ARCHIVER choices.  (I wish I could say more.)
Anyone else have that problem?

I found it...

I hit Amiga-C from the control-panel to run CONFIG.
I pick ARCHIVERS.
I pick an existing archive type.
I click the config window back/front gadget to move the window to the back,
then forward again.

The "( #3)" string is now garbage characters.

Happens 100% of the time.

--1338--

UL> FreeInter.txt               9689 ----rwed 10-Dec-93 18:39:00
UL> : Get free Internet mail-access in Detroit, MI Dec93

Who can read this file and translate it into Amiga (not IBM) terms.
And into CNet (not Tag-BBS) terms?

Here in net 1:2410 we can send UUCP mail by doing something like:
1) Address it to "UUCP, fidonet.org (1:1/31)"
   (Odd, but that's what it has to be.)
2) Feed it to my HUB at 1:2410/297.
3) Make the 1st line of the msg itself the user's uucp address.

I can do #2 and #3, but how to I do #1?

I've set-up a "NETWORK alias" like:
> uucp        UUCP, fidonet.org (1:1/31)
But "MS uucp" and "MS UUCP@" don't work.

"MS UUCP, fidonet.org (1:1/31)" doesn't work.
"MS @" won't let me enter anything as long as "UUCP, fidonet.org (1:1/31)".

Bob, (or anyone) more stupid questions about this...

How do I get BBSMENU to alias all this correctly.

BBSMENU> UUCP~MS;UUCP@1:1/31;;;;;;To:

That all works, except for the "To: ", which IS needed, but disappears.



I also have to use CNet v3.0 CONFIG/FIDONET/ROUTE-TO to force this
1:1/31 msg to go to my HUB (1:2410/297) instead.  Right?

And America-Online (AOL) users are suppose to be able to reach
me (1:2410/207) at:
> Bill.Beogelein@f207.n2410.z1.fidonet.org

My mbox-dir here is "Mail:User/1".
Or should that instead be:
> 1@f207.n2410.z1.fidonet.org
???

--1339--

Ken, please consider sticking an "idle-who" string-gadget into each
CONFIG/MODEM definition.  (Of course, SysOps could leave it empty and
"WHO" would work as it does now.)

"WHO" would just display this string as-is and show whatever info we
wanted to for its idle-lines.  We could replace the existing "Port/
UserName/Logon/Spd/From" strings, with that line's phone #, bps-rates,
public/private status, trunk-hunt order info, modem-brands, ANSI or
color-coded lines, etc.

OLD> Port User Name         Logon  Spd From                  Where
OLD> ---- ----------------- ------ --- --------------------- -----
OLD>   1    (no one)                 0                       Idle
OLD>   2    (no one)                 0                       Idle
OLD>   3    (no one)                 0                       Idle
OLD>   4    (no one)                 0                       Idle
OLD>   5    (no one)                 0                       Idle
OLD>   6    (no one)                 0                       Idle

NEW> Port User Name         Logon  Spd From                  Where
NEW> ---- ----------------- ------ --- --------------------- -----
NEW>   1  313-255-2466 Supra       144 Trunk-hunt            Idle
NEW>   2  313-255-2464 LineLink     96 Trunk-hunt            Idle
NEW>   3  313-255-2465 Zoom         24 Trunk-hunt            Idle
NEW>   4  313-xxx-xxxx HST-DS      144 Fidonet Only          Idle
NEW>   5  313-xxx-xxxx Intel       192 Paying members only   Idle
NEW>   6  313-xxx-xxxx Supra       288 Usenet dial-ins       Idle

Maybe even change the "Idle" into "Idle ## mins" for each port.
Maybe even allow MCI expansion.

A lot more informative than the just the current "(no one)  0bps  Idle"
strings.

--1340--

> address CNETREXX0
> 'transmit' "I need to use \n1 MCI codes."

Ken, could we get level #1 MCI support here?
(Or at least just "\n1" support?)

--1341--

Will someone try and get FW v2.96 (or anyone running v2.96) to do any of these
6 things:
1-2) Grab (or D/L) an unvalidated file.
3-4) Grab (or D/L) a file without a description.
5-6) Grab (or D/L) a file in a sub that you don't have D/L access to.

It's a little hard to tell how they are "suppose" to work without CNet docs,
but I would guess they would be:
1-2) NO NO
3-4) YES YES
5-6) NO NO

--1342--

BB> Yes, please do not upload QWK packets more than ONCE!
BB> I will be re-working some of the mail-handling routines to prevent this
BB> "bug" from ever surfacing in 3.0.

This will be most welcome.  Some kind of a dup-table for QWK, as well as
networking.

I hope it'll also automatically send my "carefully worded" text-file to the
QWK dup-uploader.  (SysText:DupQWKmail)

--1343--

BA> {% d}     Set the user's time remaining to 0 (hang-up) if the current
MG> So where do i do this in the ea command??

Like you said:
MG> say a week or 30 day's. i remember there was an option on the C-64 c-net
MG> if you send him a mail the had a <...> it i think he wouldn't be aloud on
MG> till his probation was up.

Just mail him:
> You can't access the system until ....
> {% d}

But maybe we could talk Ken into adding something onto each user's acct:
> Organization        :
> Sysop comment       : My SysOp Comment
> User banner         : * SysOp of this BBS, SWHQ 810-473-2020 *
> Show during log-in  :

It would be a "SysOp log-in macro" NOT a "use log-in macro".
Only the SysOp could see or edit it.
It could have MCI codes in it, and could do any 1 of 1000s of things.

--1344--

DJW> Bill, nice list, but which ones want to join-link?

Like I said in the list....

BB> Please contact me if you are a CNet BBS with 8 (or more) lines.
BB> (Or have any corrections or additions to this current list.)
BB> I'd also like to add comments about which systems specifically want
BB> to Join-Link.

I can't tell who does/doesn't welcome join-links without your help.

Please post a msg like:
> Name                : Join-Link City
> BBS Number          : 313-555-5555
> Number of lines     : 18
> Auto Join-Link days : Mon, Wed, Fri
> Auto Join-Link times: 3am-5am and 9pm-11pm

I'll gladly add it to the list.

(You can only auto-joinlink under CNet v3.0, coming 27-Dec-93.)

--1345--

I think the control-panel size-gadget should "snap" into 1-row
increments, and not allow you to re-size to 2.5 row displays, as it
does now.  (Of course, only 2 rows are displayed.)

--1346--

BB> When I run these cmds from a CLI arexx-script, MCI codes aren't
BB> expanded. Should they be?
BB> > address CNETREXX0
BB> > 'transmit' "I need to use \n1 MCI codes."

MB> CNETREXXn ports do not allow MCI. It's been that way for a LONG time. I
MB> screamed and screamed when it was made that way, and Ken told me that it

Even when I try and execute something like:
> address "CNETREXX0"
> 'bbsCommand' "TIME"

I get:

Main> \q1\n1This is your 4107th call, the system's 2199th.\n1\n1This is your
3rd call today; 77 remaining.\n1\n1Logon: Thu 16-Dec-1993  5:21p\n1Now  :  Thu
16-Dec-1993  5:23p\n2Minutes online     :     2.7\n1Remaining this call:
   391.0\n1

CNet's own MCI doesn't even get expanded.
Ugh.

--1347--

M> Is there a way to jump direct to a physical subboard?  If a sub is
M> physical sub number 300 it would be nice to be able to tell a user to go
M> direct to that sub and find the file/message.  If this is not possible it
M> would be nice to add it to C-Net 3.01!

I made this suggestion 1-2 YEARS ago.  No response from Ken.
(Not to say that it shouldn't be mentioned again.)

With some BBSes have 1000-2000 subs, it would be great.

But how would you like to see it done?

Maybe "321!" from any msg/file sub prompt, would jump directly to that sub.
          ^

--1348--
    
PB> Sorry to append this to an existing topic but the BBS will not allow a new
PB> post to be added to the system even though there is 5.1 megs available

This happens a lot on FW.  Many subs are set far too low for "max items
allowed".  I try and NOT start new items whenever possible.

I'd love to see a msg appear whenever a sub was entered:
"This sub is full."

PB> Ken the TOSS program is ignoring the NO SIGNATUE flag in the FIDO type
PB> message bases.  When I post to that area on my bbs the signature is not
PB> posted to the message on my board, but it is appended to the outgoing
PB> packet.

Ken, would there be a way to make the editor (^P preview) show Fidonet
posts E-X-A-C-T-L-Y as they will be sent out?

Complete with signatures, tear-lines, origin-lines, everything.

And after a Fidonet msg is posted, it too would show the complete msg.

I wonder how many of my callers read Fidonet msgs and see all the
signatures, tear-lines, origin-lines, etc, and then come to a Fidonet
msg that was posted locally and wonder "what BBS is this from?" and
"why are all those things missing?".
  
--1349--

JS> We have a RAD: that gets mounted in the s:User-Startup file.  It seems
JS> that when we get to the mount command, the startup aborts IF the RAD:
JS> already exists.
JS> I have tried numerous methods to get the system to HANDLE this
JS> situation, but none are working so far.

c:Assign  nil: rad: EXIST
IF WARN
   echo "Mounting new Rad:"
   Mount Rad: FROM devs:MountList
Else
   echo "Using old rad:"
Endif

(I'll add it onto your bill.)

--EOF--

Saturday 18-Dec-93 19:17:57

-Bill "Mr. BBS" Beogelein, 810-473-2020, 2-line HST 14.4k USR-DS, 1:2410/207

