
Monday 30-Aug-93 19:27:51

--1150--

Ken,
If you type "HELP DOWNLOAD" it tells you to also see "HELP EXTRACT".
If you type "HELP EXTRACT" it displays nothing.

(Also, I've always felt that EVERY reference CNet makes to the word "EXTRACT"
should really be the word "EXAMINE".  "EXTRACT" sounds like the actual files
within the archive can be extracted.  Which isn't the case.)

--1151--

BB> The only time I've seen this is when I did the ID from the Base prompt.
BB> You need to re-enter the subboard in question in order for it to recognize
BB> you as a sysop.

Ken, I've seen a lot of folks (including me) ask this question.
Is there a way to make ID's results take effect immediately?
(That is, no need to exit and re-enter subs.)

If it isn't possible or feasible, then the manual should mention "be sure to
exit and re-enter, etc".

--1152--

Ken, "FIND" boots me out of my current sub.
I was just "[v]isit"ing it.
But I think I should still be there after "FIND"ing.

--- cut here ---
> (5) Amiga JUST FOR FUN> find lex*
>
> *Upload base: CNet AMIGA SUPPORT/AREXX program files
>
>   # Upload Rsp File           Size Description
>   1 30-Aug   0 LexMess    LHA  12K This is an Arexx Script to make USE of the
>                                    LEX.LHA that Bill Allen Uploaded! Includes
>                                    the LEX.LHA and DOCS! Will Evaluate a users
>                                    POSTS, FeedBack, & Responces! :) Works for
>                                    me! :) Enjoy!
>
> BROWSE: Download, *, Examine, Grab, Quit, Read, [Pass]>
>
> List, Quit, subboard#, ?=Menu
>
> Upload base> r3
>
> Please select a subboard (L=List)
>
> List, Quit, subboard#, ?=Menu
>
> Upload base>
--- cut here ---

--1153--
It is NOT a true random-number, but the following MCI codes use the
system-time mins/secs (to the nearest 6 secs) as a seed, and will give
you a range of numbers from 0 to x-1.
(In this example x is 10.  Use any number you like.)

{L60 1400002}
{M60 #10 %}
{V60} is a number between 0 and 9.

Will someone PLEASE show me how MCI-Math Reverse Polish Notation works????
How do I do simple things like the follow with ONE {M} cmd:
(a*b)+(c*d)
(a*b)/c
((a+b)*c)/d

--1154--

> manual.  He said there is nothing IN the index under "adding commands",
> "changing commands", or anything like that.  I checked, he was right.

The index definitely needs to be expanded.
It's under "bbsmenu mods".  Page #17 of the old, white spiral manual.

Many things could help make BBSMENU MODS more intuitive, if Ken wanted to
change things:
1) Use "=" instead of "~" for cmd aliases.
2) Lift rules like "Must be added to the TOP or the BOTTOM of lists".
3) Or "Must exist in this exact order".
4) Easier methods to move cmds, especially from menu to menu, which I believe
   is currently IMPOSSIBLE.
5) Use *WORDS* like ALIAS, EQUALS, DO, RUN, ACCESS instead of cryptic , ~ | `
   chars.  Human-readable stuff.
6) Allow ";" comments everywhere and HEAVILY use them in CNet's default BBSMENU.
7) During boot-up CNet could do some mild BBSMENU error-checking.
   (Check for missing cmds, missing files, conflicting cmds, etc.)

The BBSMENU-style current in use reminds me of something that envolved
very early in CNet's life and quickly outgrew its "just list the cmds
in the right order" philosophy.

A chain-reaction occurs.  You have to change 1 cmd, to add another, but to
avoid that name-conflict, you have to change another, and another, etc.

My philosophy on writing greats docs:
1)  You have to be VERY VERY familar with the product.  But if you are, you
    overlook all the simple, "stupid" questions that a user might ask.
2)  If you are NOT familar with the product, you CAN write things from a
    new-user viewpoint and cover all the simple/stupid questions.  But since
    you aren't familar with the product, you do a poor job at it.
So it basically comes down to, "Great docs can't be done."

If Ken would have tossed the last 1000 questions he answered for users in
private-mail, into a "most commonly asked questions" base.  It would make for
some great reading.

And it would reflect some true, "real-world" problems that users face.  Things
that just can't be covered in any manual.
(But we can't forward mail from the mail-prompt into a sub #.  I wish we could.)

--1155--

Each account has a ROT13 ON or OFF flag.

Non-Cnet systems:
1)  If you have ROT13 turned on, the msg is decoded and displayed.
2)  If you have ROT13 turned off, the msg is displayed as is.

Maybe CNet could take this one step farther:
1)  If you have ROT13 turned on, the msg is decoded and displayed.
2)  If you have ROT13 turned off, the msg is skipped entirely and
    "ROT13 msg skipped" is dislayed instead.

It's pretty pointless to show full #2 msgs as:

> lkjsk kjsdk lj;lj  wekj lj slkj slkjd dlkj dfl lkjsdks dlkjfjkdlas sksk,
> slkjd dlkj dfl lkjsdk jsdk lj;ljqew  qwk ,s skdk kkk dks  fl lkjsdks d
> kj lj  dfl lkjsdk j skdk kkk dks  dfl lkjsdks dl kjsdk lj;lj  wekj l

Which is exactly what I see many times.

--1156--

Wouldn't it be great if MCI XT##### had a few more "T" values, too?
1) PortData
2) MainPort
3) UserData
4) Config1
5) SelectType
6) ItemType3
7) NewSubboardType
8) MailHeader
9) ListItem
(Or whatever the top 9 most popular structures would be.)
(The current #1-2 just aren't enough.)

A) Less 'breakage' and 'moving' around of structures each time CNet updates
   are released.
B) Less trying to count offsets of offsets of offsets of offsets.
   (This drives me crazy.)

--1157--

SD> Would an Amiga 3000t 25mhz be able to handle 5 2400 Baud Modems, a 9600
SD> and
SD> 4 14.4's?  And would about 10 Megs of Mem be anuff for this type of
SD> setup.?

You can always "simulate" memory use.
Run CNet with 10 local nodes.
Log-in locally with 10 dummy acct names.
See how much memory it uses.  (Please tell us.)
(This WON'T simulate things like ULing and DLing, etc.)

Make sure you have CNet:bbs made c:resident.

None of the above simulates CPU load, only a memory-usage estimate.

--1158--

JS> * Why is it that whenever I get the feeling I have been reading the
JS> TrapDoor Manual for 6 days straight, I look at my little counter and find
JS> I am only 17% of the way THROUGH it???

TD definitely needs a "mini quick-set-up" doc file that contains just 10 pages
of the "must know" stuff.  If you need more info, you can then read the extra
100 full pages later.

Jim, anytime you want to set-up a dummy echo between FW and my system, let me
know.  Local call.  We won't have to worry about content or dups or
"non-fidonet allowed" problems with just 2 systems in the loop.  We can both
throw in a handful of test-msgs each x days.

You can always call 313-473-2020 and DL a nodelist (sometimes I let these get
too old) under my "Gfiles/Config-files-used" menu.

JS> ** P.S.  Once I get going as a node here, does the new nodelist
JS> automatically get sent to me periodically, or will I have to go hunting
JS> for it?

I have a DOS script that I run as an event:

> Echo >>OUT:1.2410.193.0.REQ "NODELIST"       ;;; Get a new NodeList
> Mail:TD183/Bin/TrapTell CALL 1:2410/193

But that gets the FULL nodelist.
You can also just get a small "diff" file of the changes, instead.

I have a bad habit of doing #2 instead of the more correct #1:
1) Get a full nodelist once.  Apply the 'diff' file each week.
2) Get a full, new nodelist every 4-8 weeks.  Replace my old nodelist in full.

--1159--

> Alternate graph suggestions...

For showing calls, messages, downloads, the above is great.

Do you know if the above is a graph of JUST LAST TUESDAY, or an overall
average of ALL TUESDAYS?  I'd like the "JUST LAST" variety.  Always recent,
always updated, never needs to be reset to 0.

In regards to CNet's 'usage-graph'...

I always thought that the most valuable graph for multi-lined systems would be
one that showed how busy the group of 2400bps modems were VS the 14400bps
group.  (Or however they are grouped.)

Helps the sysop determine if more 2400 or 14400 modems are needed.
Helps callers to see busy/non-busy times based on the modems they actually
need/use.

(Most 2400bps shouldn't be tying up 14400 lines.  And not too many 14400 users
want to log-in at 2400bps.)

>AG 1-7           (See graph of lines 1-7.  The 2400 modems.)
>AG 8-9           (See graph of lines 8-9.  The 14.4k modems.)

Currently, CNet might show an overall 10% usage.
Even if lines 8-9 are busy 95% of the time.

(Of course, BBSMENU could be made to alias "AG 1-7" and "AG 8-9" into "Graph
Lo" and "Graph Hi" cmds, or other easy-to-remember cmds for users.)

--1160--

D> Especially when you have Carbon Copies turned on in the bases, as well as
D> the DOWNLOAD Notifications. I was watching one guy the other day; 47
D> pieces of mail. I asked him about them and he said they weren't even a
D> week old.

I think DL Notifications should be "EP" user-settable:
> Do you wish to receive DL notifications?  [No]

If 10 of your ULs each get 10 DLs, that's 100 pieces of mail.
Some people just don't want to be bothered with tons of notices.
Others do.

As now, the sysops should still be able to set DL-Notifications, but the user
should be able to override them and turn them off.

The whole story would allow:
1)  If the sysop turned a certain file's DL-Notify flag ON, the ULer could
    turn if OFF.
2)  Any caller would be able to turn of ALL DL-Notices with 1 quick "EP"
    setting.

--1161--

A> speed limitations, but if you have 24 users playing half the games on your
A> BBS, posting, downloading, chatting, and other things your CPU will be the
A> probelm not the Serail card (granted if the serial card is as fast as I
A> propose to the designers)..

I think Ken will have to eventually start supporting configured priority-task
settings.

I don't know if it should be based on users port #, his access-level,
per-user, for certain activities (games, ULs, DLs, per-pfile, etc), or
whatever.

I would gladly let the following run at a HIGHER priority for less slowdown:
1) All sysops/subop accounts
2) All paying members

I would gladly let the following run at a LOWER priority:
1) All 300-2400 callers
2) All non-paying members
3) Games that hog the CPU
4) Background message packing or searching.

--1162--

> And I fell asleep on the keyboard or something, and came to about 30
> minutes later... still in that damn editor....     I think it did it when
> entering an upload dexcription here as well...And it was giving the "Less
> that 3 minutes remain" system messages at the prompt before entering the
> edit.....

I think this is the way it was designed to work.
The clock stops while you are using the editor.
I'd like to see the clock "Slow-down x%", not stop forever.

I'm not sure what the "idle-out" timer does while in the editor.
If it doesn't already, it should idle-out just like it would outside the
editor.

(In your case, maybe idle-out was set >30 mins.)

--1163--

Ken, nothing in the v2.92 ReadMe about v2.91 or v2.92.
Nothing we need to know was fixed/added/changed significantly?

--1164--

A> rings, answers, coonects, logon, or anything) it would merely rewrite that
A> part of the who banner... I would like this to merely sitonline and watch
A> the lines...

JS>  Ken!  Look at the previous response here!!  I see I am not the ONLY one
JS> who thinks that some kind of ANSI driven online CONTROL panel would be a
JS> super cool sysop option!!

I assume you mean for this to be executed ONLY from remote.
(It would be pointless to do it from LOCAL.  Just leave your Control-Panel
open, instead.)

Take a look at my "WhoLoop.LHA".  It just clears the screen and does a WHO every
x secs.  (I know of no means by which CNet can "notify" a program that someone
new has logged in or out.  So I had to do a continuous loop.)

Since CNet is now v2.0 only, maybe Ken would impliment a method of
"notification".  Similar to the way the Amiga OS detects disk-changes,
file-changes, etc, WITHOUT continually checking for them.

I would tell my arexx or C pfile:
"Wait here until a certain task occurs"
And I could specify the task(s)...
1) A DL starts or ends.
2) An UL starts or ends.
4) A msg was just posted.
8) A new-user registered.
16) A caller logged ON or OFF.
32) A new conference-session just started.
64) etc...
(You add together the values.)

Arexx> 'WaitHere' 10

It would open up a whole new world of pfile powers.
And all WITHOUT using a single busy-loop type check.

--1165--

> 13.  + CNet echo from CLINK        14.  + CNet programming from CLINK

Hey, Echoes!  Yeah!

1)  I see the posts dated over the last month.  Have these 2 areas been online
that long?  (And my "RA" has been missing them?)  Or have the just been placed
online?  (Executing "RA" today, still missed them.)

2)  I can't seem to sort by "OR;2".  They still appear in order #1.
      0) Follow subboard default              1) Chronological
      2) Reverse chronological                3) Alphabetical
    (But only in those 2 fido subs.)

--1166--

Ken, is there any reason why 50% of the subs here give:
> Carbon copy mail not supported.

Also, since we can "CA" with ranges, I think #1 would be better than the
current #2:
1> Carbon copy mail not available in "Help me!"
1> Carbon copy mail not available in "DOS program files"
1> Carbon copy mail not available in "CNet programming"

2> Carbon copy mail not supported
2> Carbon copy mail not supported
2> Carbon copy mail not supported
2> Carbon copy mail not supported

--1167--

PB> How would you like this done,  currently if a user is online and starts a
PB> large yank, and then logs off.. the yank continues and is avail on his
PB> next call. another user could log on immediately and thus terminate the
PB> prior yank

What is the other choice?  I can't yank because the last caller started a
large yank?  I have to use up my time-per-call waiting until his yank
finishes?  This doesn't help-out by multitasking.  It just MOVES the problem
onto the next caller.

How is this handled now?

How about neither idea.  Maybe something different like:
CNet keeps track of the overall # of yanks ever run on your BBS.
Yank creates a unique dir-name based on this #.  ram:Yank1.8737
(This variable would also be an interesting value to keep an eye on over time.
"Just how popular/unpopular is yanking on my system?")

ram:Yank1.xxxx never has a conflict.  Amaint could delete all EMPTY
leftover yank dirs.  (But even if they stayed around for months,
they'd never prevent a future yank.)  And the next caller would never
have to worry if the last caller's yank is still running.  Nor ever
have to "terminate it".

--1168--

HM> Also, for some reason MCI in imported messages seems to be converted to a
HM> "\".  Any reason for this?  Why not just remove MCI on the fly if it's not
HM> allowed in a sub?

These are appearing in the CLINK echos coming into FW, too.
And should be fixed.

When should this "on the fly removal" occur?  During importing?  Check EVERY
char in every msg and remove "\xx" codes?  What will this cost in time?

Or should each user's acct have a "show/hide MCI in msgs" flag?  And CNet will
show/hide during real-time msg reading.  (Some callers love'em, some hate'em.)

Also, does "EL" need a flag that can independantly allow/prevent INBOUND vs
OUTBOUND MCI codes?  Let each SysOp control what goes out of his bases.  And
what is allowed to come in, too.  (Some echos allow MCI, some don't.  Some
SysOps love or hate MCI codes, regardless of the fact that they might be
allowed in some echos.)

--1169--

RP> it seems like more clutter. I have 4 lines and I can just fit all 4
RP> ports on one screen and I can chat with one user while watching the other
RP> 3 ports. Would I be able to do this with a 2 window chat?

It wouldn't effect your current screen/window set-up.

But then you chatted or conf'ed, you could type before, and even WHILE the
other user (or 20 users) are typing.  Your text would stay in a small separate
window.

Ever have a user that liked to:
1)  Type while you were typing?
2)  Typed too slow, and you had to wait until he was done before you could
    even START your reply?
3)  Him, or you, or both, are sitting there waiting, thinking it was the other
    person's turn to talk?

2-window chat will fix it all (and then some).

When you call in from REMOTE most terms already have this feature.  But
sitting on CNet's console, in LOCAL-mode, you don't.

--1170--

D> CONFIG capability to flag areas to bypass amaint.

I think we have a better chance at convicing Ken just to add a handful of
command-line options to amaint:
BASEKILL NOBASEKILL UDKILL NOUDKILL NEWSKILL NONEWSKILL
(See ken.lha/amaint.doc for more.)

I think he is much more likely to do the above, than to:
1) Add the above options AND
2) Write a full point&click CONFIG window to support them, too.

D> A capture Buffer in term that will allow retreving some pre-typed text and
D> transmitting it out the modem (for messages like this one)

Use a full-term (not CNet) with a ton of features like the above.

Make CNet a better *BBS*, not a better *TERM*, not a better *FAX* machine,
not a better *VOICE-ANSWERING MACHINE*, etc.  (And many of the others things
that have been suggested, which could easily double or triple CNet's size.)

--1171--

If anyone can tell me how to post a msg or a file into a sub, I'll gladly
write a C-DOS executable to do it.

Does CNet have a callback for something like:
> BOOL PostItem(struct ItemType3 MyItem, SHORT PhysicalSubNum);
Could one be added?

"File Tick/Hatching" and "monthly-fidonet-rules-posting" are just 2 of many
uses.

--1172--

I would love to get Ken to do his own CNet TrapDoor...

In the meantime, CNet could have a config-file listing
Network-session-identifier strings:
> YooHoo
> **EMSI**
> whatever
If CNet saw any of these strings at log-in, TrapDoor would be run.
(Instead of the other way around.)
95% of my calls are non-network, so that would help out immensely.

I can't get that "arexx-script answerer" to work reliably.
And I can't get anyone to tell me what *ALL* the "Network-session-identifier
strings" are that it should be looking for.
      
JS>   If the BBS is able to recognize and handle networking calls INTERNALLY,
JS> there is not need for ANY kind of "Front Door", and that is how I would
JS> like to see CNet be able to handle it.  However, there are really a HUGE
JS> number of handshaking situations out there (all of which TrapDoor seems to
JS> handle VERY well) and getting CNet to do THAT might require a long and
JS> difficult debugging process...

Would anyone be in favor of a compromise, for right now?

"recognize handshaking" instead of trying to "do the handshaking".

Have CNet answer all calls, and have CNet run TrapDoor (instead of the other
way around) only when a network call comes in.

99% of my calls are non-networking anyway.
But 100% of them have to be answered by TrapDoor, currently.

This would be so simple to do.
There are even a few small scripts written in AREXX that do it.
(But don't recognize all the different types of handshaking.)
 
--1173--

CNet currently allows us to limit the # of items in a sub.

I'd like to present Ken with a list of all imaginable alternate methods.
(Good and bad, feasible and unfeasible, sane and insane.)

If you think of any, please post here.
(The method, and reason why it would be needed.)

1)  Limit the overall # of *responses* per sub.
(To avoid subs with 3000 responses, even when you've tried to limit things by
setting "max-items" to a low minimum.

2)  Limit the # of items, as new items are posted, old items are killed.
(To avoid "Sub is full can't UL or Post" msgs.)

3)  Set independent msg VS file auto-purge-days.
(Files can take 100x as much space as msgs.)

4)  Limit posts/responses based on overall size of sub's text files.
(1000 msgs might result in 500,000 or 3,000,000 bytes of HD space.
We could keep a lid on things in real-time.)

5)  Limit ULs based on overall size of sub's files.
(Subs for Arexx-scripts use little disk space.  Subs for 24-bit pictures use
tons.  You could assign "x megs per sub", things are purged to remain within
this limit.)

6)  Kill items after x days.
    Regardless of date of last response or date of last DL.
(Old files get deleted, even if, for whatever reason, people keep DLing them.)
(Msgs only stay online x days, even if "flames" make them remain active for
months.)

7)  Available disk space.
(If there's room for another post, or another UL, allow it, regardless of any
"item-limit" settings.  I hate to turn away ULs (or even msgs) because I've
set a sub to 500-items, even if I have 146megs free on that drive today.  Free
drive space can change drastically from week to week.  No need to tweek 209
sub's "item-limits", CNet limits things in real-time.)

--1174--

Would it be worth adding a new variable into each sub's config?
> Run RR every x days:
(0=never)

If you ran "RR AUTO" it would automatically find these subs, compare
the # of days since the last "RR" and skip/rebuild as needed.

(Currently, I lose track of what subs need, don't need "RR"ing.  And
how often they need it.  And how long ago I last did it.)

--1175--

Ken, please consider allowing a sub's "scan filler" to be entered into "EL"
EXACTLY as it will be display.

Without CNet adding "()".
The user can add these himself, if he wishes.
Or use different "()" chars.
I thought "custom" was the whole purpose for this feature.

Or in my case, I've removed "()" entirely to gain some more screen
space, and can't use custom "scan fillers" at all and still align
columns correctly:

> 1.     + SWHQ: About this BBS (CNet)    2.     + Private User<-->User Msgs
> 3. (TMP) + Test Msg Base                4.     + Toss Bad or Dup Msgs Here
> 5.     + Upload QWK responses here      6.     + [RESERVED]
> 7. DIR   Amiga Local                    8. DIR   Computer Local
> 9. DIR   Amiga Fidonet                 10. DIR   Computer Fidonet
>11. DIR   Amiga Usenet                  12. DIR   Computer Usenet
>13. DIR   Non-Computer Local            14. DIR   Adult Local
>15. DIR   Non-Computer Fidonet          16. DIR   Adult Fidonet
>17. DIR   Non-Computer Usenet           18. DIR   Adult Usenet

And not being limited to 3-chars.  (That's pretty short.)
12-chars could be used for some very custom, verbose scan-filler text.
Just turning a color on and off via MCI uses up 6-chars.
Just wanting "\c1HEY\c7" is 3x as many chars as is current allowed.
We could really make things stick-out in bold/colors/inverse/underlining/etc.

--1176--

Whenever someone posts a PRIVATE msg in a BASE...

I think the CARBON-COPY that I read in my mbox should be marked
"PRIVATE" somewhere in the header.

I might treat things differently knowing whether it was sent to me
PUBLIC or PRIVATE.

And it might also help me determine if I want my reply to be PUBLIC or
PRIVATE, too.

Also, mail's header should inform the user:
> "This mail expires in x days."
I thought I had unlimited time to reply, so I held off replying to
a few of my mails.  Only to have them expire/purged without warning.

--1177--

New-user registration, less mnemonic...
> 0) Female                              1) Male
> What is your gender?

More mnemonic...
> [F]emale                               [M]ale
> What is your gender?

(and hot-key it.)

I think humans are more likely to refer to themselves as "Female" or
"Male".

I wish I had a nickle for every time I walked up to a woman and said,
"I'm a 1 and you're a 0.  Anything can happen."

--1178--

> BROWSE: [D]L, [*]mark, [E]xamine, [G]rab, [Q]uit, [R]ead, [N]ew> N

When I'm running through a ton of subs with a "NM" (New Message) or
"BA" and I see hit my favorite sub, and I'd like to read ALL the new
stuff, I'd like to just be able to hit [N]ew, instead of the current
method of listing all the shown item #s.

Typing "N" is quicker "Read 12,14,17,19-22,115,121,128....".
"N" is likely to contain fewer typos than "Read 12,14,17,19-22,115,121,128....".

--1179--

When I do a WHO I think CNet should show which users are "DLing" as
well as those "DLing and then auto-logging-off".

> Port User Name            Logon  Spd From                     Where
> ==== ==================== ====== === ======================== ================
>   3  SysOp                 7:00p 384 Livonia, MI          USA Downloading+Off
>                                      ETA:  540K @ 1703 CPS = 5 mins, 15 secs

Just in case you are trying to catch someone before they leave.

Or just in case you are wondering if that port is going to (most
likely) be free after x (ETA) mins have passed.

Also, it might be handy to see idle-times and idle-speeds:

> Port User name         Logon  Spd From                  Where
> ==== ================= ====== === ===================== ================
>  0     (no one)               144                       Idle 6 mins
>  1 m Bill Allen         5:06a  96 Livonia, MI       USA UDBase
>                                   313-473-2020 2-line 14.4k Fido 1:2410/207
>  2 m   (no one)                24                       Idle 12 mins
>  3 m   (no one)                24                       Idle 25 mins
>  4 m   (no one)                12                       Idle 17 mins

You'll know which ports are which speeds.
You'll know how long each has been in non-use.
(See which line hasn't be used in 1183 mins.  Maybe locked-up.)

Currently, all idle-speeds are 0.
And idle-times aren't shown.

--1180--

BA> I wish I had docs that told what at least 100 of the 2000 lines in
BA> BBSTEXT were used for.

D> Wait, my friend, wait! Already in the works, to be released shortly after
D> the 3.0's BBSTEXT is released. (Just to make sure lines don't change on
D> me!)

That's great news!!! Are you going to tackle 100 lines?  500?  1000?

How are you determining these?

I hope you are going to skip the obvious lines, and do the ones like:

> PBCF
> \:0\q1\n1
> .\n1
> \n1\q0\c6(\c3{v48}\c6)\c7 {v49}> \q1
> \h6ed.\n1
> \c7%c) %s\f1
> %d.%dG
> %d:\n1
> SON
> Yes
> Yes
> Yes
> Yes
> Yes
> No
> No
> No
> No
> No

Are we *really* suppose to guess at 100s of lines like the above???

--1181--

I wouldn't want it squeezed into my file-UDBASE-scans, but for msg-BASE-scans,
there is plenty of blank-space per line to add a flag-field into the display:

What are the most common 5 (or so) msg-flags?

In this proposed example, I've used:
[N]EW-RESPONSES
[K]ILL-PROTECTED  (I would REALLY like this one.
                   Save some VERY valuable msgs from being purged, when I
                   thought I had protected them, but didn't.)
[P]RIVATE
[F]ROZEN    (we could probably do without this one)
[S]YSOP-FAVORITE

>  # Posted Rsp Flags Title
> == ====== === ===== =====
> 23  8-Sep   7 N---S hooking up 2 amigas
> 26  8-Sep   6 N---- Data types
> 36  6-Sep   2 NK--- Amiga 2000 V4.x
> 59  3-Sep   1 N-P-- A2000 rev 4
> 65  1-Sep   9 N---- 1230XA mem problems
> 73 31-Aug   5 N--F- PCX Viewer
> 75 31-Aug   5 NK--S Cdb & ivs
> 76 31-Aug  18 N---- EXPLAIN PLEASE

If Ken made this "NKPFS" string, an MCI code, we could add/remove it ourselves.

--1182--

Does anyone feel that CNet should only allow users with maint-access to post
messages in networked bases, addressing it to "SysOp"?

Currently anyone can post things addressed to "SysOp" and they appear in the
private mailboxes of sysops all over the world.  Each thinking, "Why was this
msg specifically addressed to me?"

I do see those with maint-access using this as a valuable mass-mailing tool.
But it should NOT be available to the general public.

What do you think?

(And don't say, "Why was this msg specifically addressed to me?)

--1183--

Jim, in several subs here, "CA" gives "Carbon copy mail not supported.".

Any reason?  Just wondering.

Users can turn CA on/off themselves now, but only if it is allowed.
And I can't imagine why it specifically wouldn't be allowed in several of the
subs here.

An oversight?

I'm missing more and more of the things addressed to me because I can't get
them Carbon-copied into my mailbox for fast/easy "MailRead" reading.

--1184--

Is there a line in BBSTEXT that gets displayed right BEFORE the list
of subs?  So I can add a header to make it appear like all my other
menus...

> ##  DIR   Title                            ## DIR   Title
> ==  ===   ======================           == ===   ======================
>  1      + Games: Text/Help/Hints            2     + Games: Arcade/Sports
>  3      + Games: Brd/Card/Adventure         4     + Games: All Others

--1185--

Yikes....
How did my FW post get dated like this?

item: 1 of subboard CNet echo from CLINK
subj: Limiting "SysOp" addressing
from: Bill Allen  (=)
orgz: CNet Beta-tester, 313-473-2020
to  : Sysop
on  : Wed  5-May-2005  5:23a
                 ^^^^

Looks like it's going to be coming up as "new" for quite a while.

--1186--

> Documenting BBSTEXT line #s...

I can't even imagine how Ken himself keeps track of all of these.
(I assume he doesn't have any notes on them.  If he does, I sure wish
he'd release them to us.)

Of course, if BBSTEXT allowed comment-lines, you, Ken, everyone, could
document them right in BBSTEXT.
> YES       ; Line 1583, if user answers "yes" feedback
> YES       ; Line 1598, if user answers "yes" feedback at log-off
> YES       ; Line 1678, if user answers "yes" delete mail
> YES       ; Line 1783, if user answers "yes" etc, etc
> YES       ; Line 1801, if user answers "yes" etc, etc
(The comments would be stripped off when BBSTEXT is read into memory,
so it would NOT increase mem usage.)

(Those are just example lines, NOT the actual BBSTEXT line #s.)

--1187--
     
JS> few days ago (some kind of power fluctuation) and in the process the
JS> system CLOCK chip was reset to the year 2005 (May, actually).  I came down
JS> and found the system near death and reset it, but did not check the DATE!!

Probably a good check on what CNet does when a user's firstcall-date,
lastcall-date, scan-dates, post-dates, etc, end up being AFTER the (now)
current date.

Did it all work-out OK?

Also a good check of the "beyond 1999" coding.

It's only a few years away.

I hope all computers, and all software everywhere, is going to be working fine
on 01-Jan-00 (2000).
  
--1188--

I notice CNet's MOVE-ITEM cmd (MO) displays a different list of subs.
Which lines within BBSTEXT can I use to :
1)  Change the color of the sub #.
2)  Change the "(dir)" into "DIR".
for the MOVE cmd.

--1189--

Why do some headers here appear as "RECEIVED" when they aren't even
addressed to anyone...

#1/2 from: Nekron City  (Heiko Besemann)
     orgz: Cnet Pfiles Development
     on  : Fri  1-Oct-1993  6:58a  *RECEIVED*
          
How can they become "RECEIVED" by no one?

--1190--

HM>  Ken I've never had a problem getting the nodelist to open from CNet..
HM>  Could this be because these other sysops have an old traplist.library
HM>  (mine is v5.2).. ?

"Good" programs always check for a min-needed lib version #.
And report/exit as needed, if not found.

This is per CBM themselves.

--1191--

> FREQ
> Enter the address of the system for the .REQ file:
> ?

Please consider having "?" get you the same nodelist-searching prompt
as in a netmail send.

> Enter the address of the system to send netmail to (? to search Nodelist):
> Enter the address of the system for the .REQ file (? to search Nodelist):

--1192--

While Nodelist searching...
> Select a system from the list, Quit, or press RETURN to continue:

How do I quit? 
[Q] and [QUIT] do nothing at this prompt.

If your search results in a long, long list of matches, be prepared to be here
a while without [Q]uit capabilities.

--1193--

When I YANK msgs, I'd like to know which one of countless echos they
came from.  (In case I want to see if any new discussion has been
added onto that item later.  Or in case I want to add my own comments,
or even find this msg again.)

1> Item: 2 of 44 from subboard "CNet programming from CLINK"
2> Subj: Virus Checker 6.31
3> From: Andre Perusse
4> On  : Mon 27-Sep-1993  9:57a

Line #1 would be handy.

If it is too repetitive to have the subboard-name at the top of every msg,
then maybe just have it once for each sub.

1> Items 2-44 yanked from subboard "CNet programming from CLINK"...

--1194--

What is the CONFIG/FIDONET/AREA "Description" used for?

Also, in CONFIG/EXPORT-TO and CONFIG/MAIL-ROUTE, clicking "NEW"
doesn't clear the "0:0/0.0" from the string gadget.  (Like "NEW" does
elsewhere in CONFIG.)

Also handy would be....
When in the "Areas" window, "Export-To" would show a count like:
> Exporting-To    : 7 nodes
> Not Exporting-To: 4 nodes

When in the "Export-To" window, "Areas" would show a count like:
> Connected to    : 13 areas
> Not connected to: 21 areas

Sometimes these list get VERY long.

Also, there's room in the windows to make both the "Export-To" and "Areas"
List-Gadgets taller, to accommodate these longer lists better.

--1195--

I hate....
>  Respond
>  Enter the ADDRESSEE'S handle [No one].
>  : Dotaron              .
>  Not a user of this system--use anyway [No]?

I'd love the CNet standard...
>  Enter the ADDRESSEE'S handle [No one].
>  : Dotaron
>  1.  Don Powers
>  2.  Don2
>  3.  Doomfane
>  4.  Dotoran
>  5.  Doug Lang
>  6.  Doug Man
>  Enter number, name, +, -, or [RETURN]: 4

At least for non-networked msg bases.

--1196--

We need to get Ken and a handful of others to:
1)  Make a DOS script containing a random selection of c:date, c:dir, c:list,
    and other "safe, working" standard DOS cmds. 
2)  Run in on your BBSTEXT "@logoff complete" line via MCI.
3)  See if you get lock-ups at log-off a few times during the next 100
    callers.

#2 is getting to be a pretty popular place for a wide variety of pfiles.  
It is NOT some place we need a "hanging" problem in CNet v3.0.

--1197--

> FW's clock accidently set to the year 2003...

Jim, here's another "Wed  5-May-2005" post that needs to have its date fixed.

("BG SINCE 10-10-93" should find'em all.)

But...

What would have happened if amaint would have been run while the system
clock was set at 2003?

All files, all items would be shown as not having been accessed in years,
and would be purged.  A pretty expensive mistake.

Amaint definitely needs some precautionary date-range stuff added.

Maybe have it track/check the current-date and the last-amaint-date.
If these dates are +/-30 days apart, just abort with an error 
msg (displayed and logged).
(It would be unlikely to not run amaint in 31 days, or whatever.)

--1198--

> Enter the address of the system to send netmail to (? to search Nodelist):
> 1.  Address:  1:3823/70

Ken, are you using calls to TrapList's FindNode() function?

It takes forever to search for a node like 1:3823/70 (near the bottom
of the list).

I realize it is searching through 1.5 megs of data, but...

170 secs for the 1st match.
310 secs for completion.

My NL_REXX.LHA node searcher, written in (slow) arexx finds it in 3
seconds, with the help of a FindNode() call.

Also, I feel that "Set which pattern, or press RETURN to begin:" should
be hot-keyed like the similar prompt for "UL" (User Lister).

(BBS-Lister needs this hot-keying action also.)

--1199--
This is with CNet v2.94...

> Of the 4,747 files online, new files appear in 2 area(s).
> 
> Quick-browse for new files [Yes]? Yes
> 
> *** File Base (root): Telecommunications/BBSs/Telcom: Text & BBS # Lists
> 
>   # U/L'ed DLs Filename   EXT Size Description
>  == ====== === ============== ==== ===========
>   1  3-Oct   0 WABL9310   LHA  20K Worldwide Amiga BBS List (WABL) v93.10 02Oct93
> 
> BROWSE: [D]ownload, [*]mark, [E]xamine, [G]rab, [Q]uit, [R]ead> 
> 
> *** File Base (root): Quick U/L (to any area)
> 
> *** File Base (root): Quick U/L (to any area)
> No items matched.

I regularly move U/Ls from "Quick U/L (to any area)" into their correct
subs.  That's what I did above.  CNet still thinks there are new files in the
U/L sub.  (And it displays the sub-name twice.)

I guess I don't mind the subname appearing, after all, there WAS (past
tense) a new file there.  But could all the double sub-names be fixed?

--EOF--

Monday 04-Oct-93 16:12:17

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

