
Wednesday 05-Jan-94 06:01:42

--1400--

To Ignus Fast

Here are most of the cmds in BBSMENU.
Would anyone care to put together SysText:Help files for them?
(Following the same "Command:/Purpose:/Available at:" format as Ken's.)

CNet has 40 help files out of the possible 160+ cmds.

+ - . Abort ACcount Add AG Again AL ALPhabetize AM AO ATtribute BA
Base BG BN Border Browse Cancel CArbon CC CD Center Chat Continue
COntrol Copy CRedits DD Delete DELete DIR Download DRop DS DUmp EA
Edit EF EG EL ENtry EP ET EU Examine EXit Expand Feedback Filter Find
FInger FOrward FReq Get Gfiles Grab Group Help HIde IDentify Indent
INFo Insert Invite Join Justify Keyword Kill KL LA Last LC Left List
LOGOFF LU Mail Mciread MEmbers ML MM MOnitor MOve MR MS MUffle MV NAme
New News Next NF NM Noreceipt NOTE NSal NU O Off OLm ORder Over Pack
Pass PErmanent Pfiles POst Put PW Quit Quote QWK RA Read Relogon Reply
Respond RF RG Right RM RN Room RR RUN SA Save Scan SEArch SElect SG
Shell Skip SN SS STatus Step Subboards TErm TEst Time Toggle TRansform
TYPE UI UL UM Unindent Uninvite Upload Uploads Validate Verify VF VIew
Visit Visual VN VOte WF WHo WHY Wrap Write X Xamine Xit XR YA Yank YG
YN Z ZG Zip

--1401--

> #1/1 from: Bill Allen
>      to  : Anubis
>      on  : Sat  5-Jun-1993  9:49p  *RECEIVED*
>
> Find *.ZOO
>
> Try [E]xamining "trump.ZOO" (and many other) .ZOO files.
> Can't be done.

This still doesn't work in CNet v3.04.  About 90% of the .ZOO files
can't be viewed.  Anybody know why?

--1402--

RE> I think Ken should put in an "attach file to message" feature such as the
RE> one in Galacticomm's MajorBBS.

There would be a bazillion things to configure...
Which partition(s) would the files go to?
How long do they stay there before deleting?
What happens after they are D/Led?
Are they automatically deleted, let the sender decide, let they receiver
decide?
What the total # of files that might be awaiting all your users?
How do you limit that?
What are users charge for trading files?
Which access groups can U/L, D/L, etc.
Should the U/Ls be tested, transformed, etc?

All of the above, and more, is already handled if you have a "Private
U/L" sub setup.

--1403--

So we went from having:
1) Two different versions marked as v3.03.
   The real v3.03 and a true v3.04 release.

to:
1) Two different versions marked as v3.03.
   The real v3.03 and a true v3.04 release.
2) Now a second v3.04, but different the eariler v3.04 (marked as
   v3.03), but also is still marked v3.04.

Anybody can forget to bump-up a version #.
When it happens, don't make it worse.
Just make the next release +1, and forget it.



Also, "Version CNet:BBS FULL" should have the date in it.
                        ^^^^

SASC> #define CNET_VERSION "$VER: CNet v3.05 " __DATE__
CLI > Version CNet:BBS FULL                (^ space)
CNET> CNet v3.05 Jan 12 1994

It's NOT the (weird) date format that CBM wants everyone to use
(12.1.94 ugh), but it's a lot more readable, and 100% automatic every
time you re-compile.  (Even if you forget to bump a version #.)

(Picky, picky, picky.  Everybody's a critic.)

--1404--

Since I can't just overwrite my BBSconfig3 file with the updated version
each time, (without losing tons of my config data), how do I update things
like the new CONFIG/LOG names?

--1405--

> Hey, My Userdata is garbled, do you have any programs to restore, salvage
> or anything for this ? Please I'm In DIRE trouble..

Nope, sorry.  Please post these kinds of things in a PUBLIC local or Fidonet
sub, not PRIVATE mail.

1000 SysOps can help far better than 1 (me).

Also, I backup my entire SysData: dir into a dir called MON each Monday, a DIR
called TUE each Tuesday, WED each Wednesday.

(Automatic system event each midnight.)

7 copies of back-ups without lifting a finger.
And the old (maybe corrupt) data can't overwrite the new (non-corrupt) data
for at least 1 week.  (Which I would have noticed long before then.)

--1406--

Is anyone having any luck KILLing vote topics?

CNet tries to do #1, instead of #2, when killing vote #31:
1> delete >nil: sysdata:vote31 all
2> delete >nil: sysdata:vote/31 all
                            ^

Won't that cause all the killed vote topics stay on the HD (but not
shown online) forever?

(Ken did you see my post about this several weeks ago?)

--1407--

I just D/Led this file and the file-DL-count stays at 0.

> item: 1 of subboard AREXX Program Files
> file: CNet_Math.LHA (6952 bytes, 0m 4s, 0 xfers, 0 cps best)
> from: Bill Allen  (=)
> orgz: CNet V3.01 Beta-tester
> to  : Rinaldo Petterino  (=)
> on  : Mon 10-Jan-1994  2:44p  *RECEIVED*
> info: New CNet-BBS cmd computes math equations online
>  (NOT YET TESTED)

Maybe because it's "Not tested"?
Maybe because it's addressed to 1 person?

--1408--

PB> Bill clean out your mailbox... Tried to answer E-Mail first so not to
PB> continue this convo in public.

It just gets filled as fast as I can clean it out.  Please keep 99% of
everything PUBLIC.

PB> BA> PB> Welp, there's going to be HELL to pay.
PB> BA> What is "hell" charging these days?
PB> In my case, it's tedious explanations that we know these bugs are here,

Use PUBLIC posts for 99% of the stuff.

I too answer the same questions over and over and over in private.  "Keep it
public and 1000 can read it.  Make it private and 1 person can read it."


Maintaining a single master list of known bugs sure would help.  Anyone could
type "BUGS" at any prompt and see:

Bug Descriptions   | Found by |  Date   | Vers. | Date fixed | Vers. fixed
Wrong file marking | ID #999  | 01Jan94 | v3.04 |   ??Jan94  | v3.0x
                   |          |         |       |            |
                   |          |         |       |            |
                   |          |         |       |            |
                   |          |         |       |            |
                   |          |         |       |            |
                   |          |         |       |            |

It would instantly answer 8 of the most often asked questions about bugs.
> Where are there?
> Which have already be reported to Ken?
> Who's having this problem?
> When was it found?
> Which version is it in?
> Has Ken already found and fixed it?
> How long has it been around before getting fixed?
> What version fixes it?

Well worth 1 line of text per bug.

--1409--

I wouldn't think that thanking users by HANDLE looks very UNprofessional in a
commercial package.  "We want to thank HellHole, and BloodSword for their
great work."

But I do think that thank-yous (mentioning real-names) should have been part
of the manual.  Was it?  If not, maybe just an accidental oversight?

--1410--

I sure wish ALL beta-tester would identify themselves with CNet's handy
"organization" string, "who" string, "signature" file, or SOMETHING.

(With or without mentioning the version # that they are currently running.)

--1411--

Hey, there's an interesting date on FW running v3.04...

> Enter item number(s) or title(s) to Scan (?=help):
>  : all
>
> # Upload Rsp File           Size Description
> 1 110-Ja   2 CNet_H&S_v2LHA 204K This is v2.1 UPDATED Hack & Slash DOS
>                                  Pfile.
>                                  It has a new Arexx loader that sends the
>                                  UID# instead of the Handle of the player
>                                  and comes with CNET INSTALLATION
>                                  instructions written by Dotoran!
>

The 110th of January?

--1412--

P> Well I would prefer it to the method in the manul, w/o the @ symbol.

I too would like the "NO @" method.

P> But that could be confusing, MS jim could match two users, on who's
P> handle may be "jim" and one in the alias file.

It would be a simple matter of picking #1 or #2.  If I SPECIFICALLY set "jim"
to an alias that I want, I would think that would be the 1 I want.  No?

Or at least make that 1 the default, or maybe the #1 choice.

> MS jim
> 1.  Jim Selleck@1:2410/215
> 2.  Jim Johnson
> 3.  Jimmy Dean
> 4.  Jim Jones
> 5.  Jimmy Carter
> 6.  J Smith
>
> Enter number, name, +, -, or [RETURN]:

D> As far as I know, if you include a space, to me that's saying "Send mail
D> to the user whose first name is Ken and whose last name is @",

That must have been changed in v3.04.  In v3.01 is DEFINITELY allows MS KEN@
or MS KEN [SPACE] @, and does the same thing.  (As it should.)

D> 1: Is there a Ken Pletzer on THIS board? If not, go to number 2, but if so
D> 2: Check the alias file for a Ken Pletzer and use that info in a second
D> 3: Present to the user the Handle/Real name listing.

I say do #2 before #1.  If the user SPECIFICALLY set-up an alias, chances are,
he means to use it.  (And use it often.  I thought that was the whole idea
behind aliases.  "Shortened forms of frequently used addresses.")
                                     ^^^^^^^^^^

D> Main> ms Dave
D> Send Mail to Dav? [Yes]: No
D> Send mail to David Weeks @ 1:260/121.0? [Yes]: No
D> 1. Dave
D> 2. Dave Czaya
D> 3. Doug

Why make him answer 2 questions AND the answer the userlist question?
(We are already questioned-to-death just to send mail.
Subject/expire-days/filemail/receipt/Return original/crash/Urgent/etc.
Don't add 3 MORE questions onto the list.)

Why not just the userlist first (and only) if there's a name conflict?
(And you won't even need that, if there isn't a name conflict.)

BA> > 1.  Jim Selleck@1:2410/215
BA> > 2.  Jim Johnson
BA> > 3.  Jimmy Dean
BA> > 4.  Jim Jones
BA> > 5.  Jimmy Carter
BA> > 6.  J Smith

Your aliased-choice will always be the #1 item, so you can just hit "1"
without even thinking about it.  Of course, "MS jim!' will even do that for
you.  (Assume your aliased-choice.)                ^
                                                   ^

(All of the above are FUTURE IDEAS, not current working methods.)

"Jim Selleck" sure is a popular name around these parts, isn't it?

--1413--

I would like to see 2 changes to the conferencing area...

A) If someone enters a conf-room that has the "Chaos" flag ON, then have
his WHO/DOING string say "Main" instead of saying "Conference".
(Otherwise, it's pretty easy to tell "/WHO" you're chatting with.)

B) Regarding conf's "13) Lurk timeout (min):"...
Who should be exempt from it?

1 Those with sysop-access?
2 Those with control of that room?
3 Only those with 1 AND 2?
4 All of the above?
5 None of the above?

I'd like to see #3 (or maybe #1).  (Currently it's #5)

--1414--

M> I was wondering if someone could upload a sample script using the CNetToss
M> and Fido/clink.  I am not to good at writing scripts yet and could use a

Here's my FULL CNettosser DOS-script:

> ChangeTaskPri -1                       ;;; Don't hog the CPU
> Run rx CNet:Rexx/SysOLM.rexx K         ;;; Send OLM to everyone online
> CNet:Fido/Toss VERBOSITY 3             ;;; Toss it

Exactly what do you want your script to do?

--1415--

PB> because of the special serial numbers we had been assigned.   This was to
PB> be our "Badge of Honor" for devoting so much time to the development of
PB> this Fine piece of software.

It's the person with the *HIGHEST* serial # that is the better man, gets more
dates, and has fewer bathroom mishaps.

--1416--

I really love "3) Message reading : ANSI scrolling".
Any plans on allowing that setting to also effect mailbox reading?

--1417--

P> Has anyone got a LARGE list of BBS's around the country (in Cnet format)
P> that they could upload?  I'd sure appreciate it!  Mine got trashed *8^<

FW has the largest list (in CNet-format) that I've seen.

Ed Dukeshire's WABL is the largest list (in text format) that I've seen.
(Type "FIND WABL*")

Jim, please type "LZ U t:CNetList.LHA SysData:bbslist" and move t:CNetList.LHA
into a sub some place.

I too would like love a copy.

Thanks

--1418--

TB> I wrote the original Installation script for CNet 3, and can say that at
TB> least it USED to deal with ALL of the PFiles:BBS/ executables.

I don't have the installer script, but I've been see many OTHERS say there are
problems with it.  If so...

Bob, no offense, but the installer script should have been beta-tested for
just as many weeks as CNet "should" have been.

In addition to finding bugs (if any), making suggestions (if any), all the
beta-testers would have had many weeks of a VERY useful installer-script.

Am I correct in assuming it got 0 days of beta-testing by other users?

--1419--

SD> This may seem like an Off the point.. However, the FIFO-Handler sent with
SD> the C-Net disks was Auful.. locked my system up.. so I went back to Matt
SD> Dillions.. have not had a problem sence.

Which 1 was that?

fifo-handler        7972 ----rwed 06-Jun-92 20:44:26
fifo-handler        6948 ----rwed 10-Jun-93 02:36:10
fifo-handler        6960 ----rwed 21-May-93 01:36:45
fifo-handler        7032 ----rwed 02-Jul-93 14:10:55
fifo-handler        7024 ----rwed 28-Dec-93 00:27:08

Ken WON'T put version #s in any of them!!!!

--1420--

RW> port I notice it is sucking up approximately 1.8 megs of ram just to load
RW> a port. I have an A3000 with 10 megs of ram, and suddenly cannot even run
RW> the BBS anymore.  What on earth could be causing this?  I have the stack

Are you making CNet "resident"?

If you EVER run more than 1 port (counting your local port) you should
DEFINITELY make CNet resident.  There's no reason not to.  Why load and run
2-20 different copies of the same program (CNet:bbs) instead of just 1?

It's just 1 simple DOS cmd...
> Resident CNet:bbs            CNet:bbs           PURE

I order to check that you've done this correctly...
> Load 2 ports.
> Type "RESIDENT" from the CLI.
> Do you see:  NAME              USE COUNT
>              CNet:bbs              2


Also,

RW> --- CNet XFIDO 2.41

Upgrade to v2.63 for free.
Or upgrade to v3.04 for a fee.

--1421--

> MCI_Says logs where?

DW> 1: Using the LOGENTRY ARexx command(or the {+ } MCI Command)

The LOGENTRY cmd gives "arexx_says".  (As it should.)

The {+} cmd gives "mci_says".  Thanks.

DW> 2: Using 1 or more "SysOp Only" MCI commands within the program, where for
DW> instance any uses of the {$ } command might be logged.

I can't get {$} to log anything.  Pretty dangerous.  If there's hidden {$} cmd
in a text-file or arexx-script, I DEFINITELY want them logged.

Can someone else try and get {$} to log something?

--1422--

> "Sex: [M]ale, [F]emale?"
> "Keep your sex private?"
would also be a very PC (politically correct) change.

--1423--

S>  Is there any plans to allow CNet to run on 2 different computers
S>  and still be able to fully communticate back and forth like many
S>   of the IBM BBS' can, via ether net cards?

I'd like to see a serial<->serial connection instead.
(Or can that already be done?)
I'd rather buy a $5 cable, than 2 full ethernet cards, wouldn't you?
And I can't imagine ethernet being faster than a direct serial<->serial
connection.  Is it?

(Actually, I guess a par<->par connection would be better.  Save those
precious, very few serial ports for phone line use.)

If people are inquiring about this so that they can run >20 lines, are there
really THAT MANY people that want to run 30-100 lines?

--1424--

On Fri 14-Jan-1994  8:28a, Big Brother wrote:
BB> The DMS disks that you download from this subboard come equipped with a
BB> "DEMO" bbslicense file.  These DMS disks may be freely distributed as
BB> downloaded.  When installing these DMS disks, be careful not to overwrite
BB> your registered "bbslicense" file with the "DEMO" version.

On Mon 17-Jan-1994 12:25p, Jim Selleck said to Bill Allen:
JS>   I'd like to do that, but have to check with Ken first.

I thought that WAS Ken's OK as of last week.  No?

"may... be... freely... distributed..."

JS> Also there is a
JS> LITTLE matter of moving the location of that sub so there is enough SPACE
JS> for those archives!

I know all too well about HD space.  Hopefully you'll have time to juggle a
few things around, as needed.

After 2 YEARS of waiting for a demo, the 2 last stumbling blocks, now that
it's 100% available are:
1) Does Ken mean "freely distributed" when he specifically
   says "freely distributed"?
2) Disk space on FW.

I have a bazillion people to give that demo to.  It's been in the "Jim Only"
sub since last Friday and I've seen several requests by others for it also.

--1425--

PB> Ken addressed this in E-Mail a while back and told me that it would be
PB> added, but, not till after the 3.0 release. and it would be sped up if you
PB> could paint a complete scenario for all options

The complete scenario would be exactly the same as ULing and DLing.
Just ask the user...

> Use your default protocol Zmodem [Yes]? No
> A) ASCII
> X) Xmodem
> Y) Ymodem
> G) Ymodem-G
> Z) Zmodem

I can't imagine what else to add.

--1426--

A> and eliminate stupid Id
A> numbers so any system can call and link in without screwing up numbers and
A> crap, just use bbs name or a assign-when-you-call system....

Yes!  I've mentioned to Ken, "ID #s should be unnecessary" and "passwords
should be optional".

If CNet needs an ID #, have it use a # computed out of the phone # called, or
the user's CNet serial #, or any kind of a unique per-system #.

Or just "use the next ID # available".  If systems #1, #2, #3, and #4 are
already joinlink, the next becomes #5 automatically.  Never any conflict of
any kind.

And passwords should be optional.  If you set your JL up without one, everyone
will be welcome.  No more "Can't joinlink, passwords don't match".  No more
need to keep a list of all the passwords of all the joinlinkable systems
around the country.  (And the frequent changes.)

Make JL and easy and usable and freely open to everyone, as possible.
(Of course, you could still use a password if you wanted to run a closed JL.)

--1427--

S> I think you can assume almost any arexx program run through the updater
S> will work fine, any dos program should work fine, and any older CNet C
S> program wont work at all :)

Which arexx updater?  The 1 that changes MCI codes from their old \ to {}
methods?  Or is there one that changes all the old XT##### codes into their
entirely new locations?  I can't imagine how an old arexx XT#### pfile could
work otherwise.



Also, I think a GIANT message should state....

"If you experience problems after converting to v3.04, shut-down your
pfile-menu for 1 week, and remove any log-in/out pfiles, and see if they go
away."

(One week without pfiles won't kill anybody.)

--1428--

RE> As far as I could tell, NO, there is no way to tell the difference.  Maybe
RE> Ken should enhance getuser 17.

Definitely!

Instead of the odd, cryptic "0" VS "1234567890" VS "impossible", make getuser
17 just return human-readable "NONE", "PARTIAL", and "FULL".

Or is it too late to do this?    :-(

--1429--

RE> Bill, we are going to be holding a CNet (voice) Teleconference on
RE> Saturday, January 22, 1994 at 2 PM (EST) with Ken Pletzer as the "host".

Yikes!  Short notice.
I just got this mail Friday night.  I won't be available at all Saturday
afternoon.

Why voice????

We already have a 6 user BBS for confs.  (FW)

And we can joinlink in 10 more 8-22 line systems!!!!

We could have all 30-40 beta-testers online without ANY kind of special AT&T
voice-Teleconference setup.

And everyone is far, far more likely to talk openly in text, instead of voice.

And everyone's words won't be jumbled together while trying to talk at the
same time.  And everything can be captured to a file for later review, and,
and, and....

--1430--

I would really like to see CNetToss have independant "TOSS" and
"BUNDLE" options.

If I have 2megs of msgs in InBound: and just 2000 BYTES of msgs in
OutBound:, I might only want to process the Outbound. right now.
(Due to time, speed, diskspace, CPU load, etc.)

Or I might only want to process the Inbound, waiting to bundle my
outbound stuff until after I take a peek at it.

--1431--

If I limit access to BBSMENU cmd with:
> FINGER `23-31

Online, CNet reports:
> Main: FI Ken
> Unknown command "Ken" ... enter ? for a list of commands.
                   ^^^

The cmd was "FI" not "Ken".

--1432--

D> fb : feedback
D> ms : mail sent
D> m1 : mail sent to ID # 1
D> po : posts
D> re : responses
D> gf : gfiles
D> ne : news
D> uf : uploaded files
D> uk : uploaded kilobytes
D> df : downloaded files
D> dk : downloaded kilobytes (you forgot this one)
D> mu : minutes used
D> mi : minutes idle
D> ch : number of chats

1> fbmsm1poregfneufukdfmumich
2> fb: ms: m1: po: re: gf: ne: uf: uk: df: mu: mi: ch:

BB> Dan, if Ken had used #2 in BBStext, instead of #1, would it have
BB> made it clear without asking?

Ken, please consider adding a "$$:" string for "total charges for that
call".  ("Rounded to the nearest whole cent" would be great.)

--1433--

TB> XXXXX has no description.  To give it one, go to the subboard you uploaded
TB> it to, and type W 'XXXXX'.
TB> SO, what I need is to be able to spit the filename out twice.

He's how I'd like to see it work...

Why make the user go hunt/find the correct subboard (and sub/sub/sub/subboard)
and hunt/find the correct file.  A file description is important.  Just ask
him for the filenote right at log-in...

> FileName.LHA needs a description.
> Enter one now [Yes]:

Bug the user on EVERY call until he enters one.
Or he can keep hitting "NO" forever.  (If he feels that's more productive.)
:-)

But I bet you'll get all your files filenoted a lot faster (and easier) than
any kind of "go find the subboard, go find the file yourself..." msgs will do.

--1434--

Ken, I just got 154 errors during an U/L on port #3 here.
510K file, 14.4K bps.
I don't know if the problem is on my end or yours.
There are 5 users online.

Also, please have CNet write "the # of errors occurring during each file
transfer" to the logs.  Thanks.

--1435--

If I have a few pieces of mail that I definitely want to save, how do
I forbid someone from going into my mailbox and EDITING/KILLING them
(MV) just because they were the original author?

I was hoping that if I've already read the mail, MV-EDIT/KILL would no
longer be possible.  Once I read my mail, it should become MY private
property.

I can currently send Ken a msg, let him read it, let him specifically save
it for future reference, and I can STILL go back later and change it, and
then claim that I said exactly the opposite of what I had originally said.

"MV" is a valuable cmd, so I don't want to have to remove EVERYONE'S
access to it, just because of that.

Here's how I think it SHOULD work....
If the receiver HASN'T read his mail, the sender CAN   MV-EDIT/KILL.
If the receiver HAS    read his mail, the sender CAN'T MV-EDIT/KILL.

(Of course, sysop-access would still allow you to do either one.)

D> This wouldn't be too hard if Ken made it so it simply changed the AUTHOR
D> of the mail, since I believe the MV command is checking either the author
D> or the serial id number of the sender. I believe it may be the latter. If

I would rather NOT have CNet mess with altering author names or serial #s.
(They might be needed for replies, etc.)

Just have CNet check the (already present) *RECEIVED* flag.
If it's been *RECEIVED* you can't edit or kill it.

Does anybody see any horrible things that might happen if this was done?

--1436--

> 0.  Default (others override)
> 1.  Local mail and feedback
> 2.  UUCP messages
> 3.  FIDO messages/handles
> 4.  FIDO messages/real names
> 5.  Other messages/handles
> 6.  Other messages/real names
> 7.  Upload descriptions/handles
> 8.  Upload descriptions/real names

Is a separate signature-file planned for "all netmail sent" in CNet v3.1?
(Regardless of the network used, in multi-network set-ups.)

CONFIG/FIDONET/NETMAIL-ORIGIN is great for a short origin-line, but it
can't hold a full 2-4 line netmail signature.  And, of course, isn't
user-dependent.

A>  Is a separate signature file planned for all "anonymous" posts?

heh, heh...

An anonymous signature???

Actually, I guess someone could have some kind of a need for unique, odd,
"signature" file for all his anonymous posts.

--1437--

BB> WHERE do you set this?  I tried changing C.A so that __STACK was set to
BB> 40000. I then assembled C.A, compiled control, set the stack to 10000, ran
BB> control, did a STATUS FULL, and it still reports 10000, not 40000.
BB> What am I doing wrong?

You are using "__stack", and not "__STACK", right?   (lowercase)

long __stack = 25000L;
main()
{
}

(P.S. I've never tried any of that.)

I posted the following msg in the SAS/C Fidonet echo.
(FW really should be picking up this great question/answer echo.  
Tagname "AMIGA_LC".  "AMIGA_PROG" is also very good.)

BA> /*
BA> 
BA> When I run the following code, c:Stack will still report the stack
BA> setting for that task as the same for the CLI that it is run it.
BA> 
BA> Am I doing something wrong?  Is c:Stack wrong?  Or will it never report
BA> 25k stack for that task?
BA> 
BA> Should I be linking with something special?  Using certain start-up code?
BA> 
BA> */
BA> 
BA> long __stack = 25000L;
BA> void main(void)
BA> {
BA>    char buf[99];
BA>    gets(buf);
BA> }

--1438--

JS> For anyone interested: I learned at the Fido region meeting last night
JS> that SOON we will be able to easily and quickly gate Internet-netmail
JS> through my Fido boss (Gary Gilmore) who has an Internet account and
JS> polls several times every day.  ALSO, a local feed for UseNet
JS> newsgroups will SOON be available!  So by March I should be a much
JS> better educated CNet SysOp!!!!!

Jim, this is great news.  I've been using the small arexx script
"UUUCPgate.LHA" to send Internet-netmail via Fidonet from 1:2410/*.

But really looking forward to a local feed for actual newsgroups.

(I didn't know you were going to that meeting.)

--1439--

I think the confusion comes in when we have a s:startup-sequence file
call a 2nd script to run the control-panel which in turn, runs CNet
which in turn, runs VDEs and yank-tasks, etc, and events, and displays
text files, which also in turn, call more scripts and more
executables, etc and etc...

It gets very easy to lose track of which prg inherits which stack from
where.

I'm all for having:
1) All CNet executables set their own stack.
It's a shame to run CNet with an 80,000 stack, and then see it call prgs
that could easily run with 1K of stack, but are run with 80,000 because
that's what they inherit from CNet.
(Think of what a 10-20 line system is like.)

2) Each CONFIG/EVENT and each pfile-menu VDE have a "stack" gadget.
   (For 3rd party stuff doesn't support #1.)
(Same reason as above for #1.)

TB> The only problem with that is for programs which you run from the shell
TB> which to NOT require such high stack, you are wasting extra memory.

TB> I kind of get a kick out of the people running 100k stacks on CNet, since
TB> they're probably blowing at least that much for every 2 ports they have
TB> up.

That sums up the problem nicely.  Now, what's the solution?

Keep in mind, we are going to have programs that call programs that call
programs, etc.  (And automatic events run.)  And every single one is going to
use a different amount of stack.

So I think users feel (right or wrong) "set it extremely high, just to be
safe, for the 1 prg that might need it".

Personally, I'm not sure why the Amiga even needs to distinguish between STACK
memory and non-STACK memory.  (Anyone know?)  Why not use exactly whatever is
needed for variables and non-variables?  No more no less.
                                                              
It's kind of wacky to tell users, "Here's a great new program.  It needs an
unknown amount of memory for variables.  Make sure you set your stack high
enough to allow for that amount."

(But that's exactly what EVERY Amiga program does.)
                         ^^^^^

--1440--

P> Thanks!! I'll fix it as soon as I get my 3.0!! As I think the problem lies
P> in how my system vs. other systems, interprets DOS doors, 

How would your system differ in how it interprets DOS doors?

Also, it would be priceless if your revision history stated:
> Tested with CNet vX.XX
for each of your updates.

I don't know which of the many "we don't know what version it works with"
pre-v3.0 and post-v3.0 pfiles that are going around these days, are worth
trying under CNet v3.04.

Pete, arc-viewing is a VERY handy feature.  Sure wish we could convince Ken it
should be a part of CNet itself.

CONFIG/ARCHIVERS already has all the necessary view/text/pack/extract formats.

We would just need a BBSMENU arc-prompt that allowed:
> Read    c:Type %s
> Version c:Version %s FULL FILE
> Whatis  c:WhatIs %s
> Pack    s:RePack %s

After CNet prompted the user for the filename, it would be extracted and he
could Read/Version/Whatis/etc and 100 other powerful things.

And it would be available on ever CNet in the world, NOT just the very few
that D/L and install FPAV110.LHA.
  
--1441--


If you are not interested in this subject matter, please hit the
space-bar and skip this msg....

If you call a BBS and read a msg posted by "anonymous" that contains
ABSOLUTELY NO name-calling, illegal material, or deliberate wrong
information.  Just the usual assortment of facts, opinions, comments,
and questions that any user posts under his real-name or his handle...

What goes through your mind?

I hate that guy's guts for posting "anonymous" things!
I can't stand it when some one posts something "anonymous"!
Posting by handle "Bottle-Fly" is much more informative than "anonymous".
I will go to any lengths to find out who it is!
Maybe I'll just start accusing anyone online at the time!
   (Regardless of HIDE'ing or non-HIDE'ing callers.)
I will *NEVER* read anything post by "anonymous" again!
I absolutely refuse to reply to all "anonymous" posts!
I think this sysop should forbid all "anonymous" posts by everyone!
I'm totally confused by "anonymous" posts!!!
"Anonymous" posts use much more diskspace than non-"anonymous" posts. (???)
"Anonymous" people should be refused all help, and all support.
"Anonymous" people should be locked out of the system.

(When responding to this item, please also state your age.)

Here's my response.

As long as these conditions are met...
> If you call a BBS and read a msg posted by "anonymous" that contains
> ABSOLUTELY NO name-calling, illegal material, or deliberate wrong
> information.  Just the usual assortment of facts, opinions, comments,
> and questions that any user posts under his real-name or his handle...

I could care less who "anonymous" is.
He should not be hated, ignored, avoided, skipped, deleted, or locked-
out just because I see the "anonymous" word in the header.

He doesn't offend me, shock me, harm me, or infuriate me in any way.

I'm "over 21" and have had the same attitude (as shown above) since I
was about 6.  But I see a lot of very hot-headed people feel otherwise here.

--1442--

> export-to during a toss.  Set this to 0 to disable to feature.  To
> have this feature work, you need to create a directory
> mail:dupetables/ where the actual data will be stored.

Would it help speed things up if we had a CONFIG/PATH for the
dupetables too?

My FFS RAD:drive is 2x-3x as fast as my harddrive.

(And no HD fragmenting, or validation-errors during crashes, to worry
about either.)

--1443--

This would cover it nicely...

JS> Subboard Full: No new items
JS>                Delete oldest
JS>                Auto-expand

I too would use msg-wrap-type deleting.

If my feed has a problem, and no msgs come in for a few days, I now watch my
subs dwindle.  What originally was a VERY busy sub, that I SPECIFICALLY wanted
to keep 1000 msgs in, then becomes 500, then 100, then 20, etc.

So it's either "swamped with posts" or "almost empty".
(Who wants an extremely busy sub that's almost empty?)

Wrap-deletion would keep the # of posts at n, AT ALL TIMES.


Even without any feed-problems, Wrap-deletion would be handy.
No need to guess what kind of light/heavy activity might occur in a new sub.
Just set the # of items to n, and that's it.  n == n at all times.

--1444--
BC> S> I dont think you can.... you MIGHT however try to set the persons
BC> screen
BC> S> height to some god-awful number... its getuser 1100468 -- make you sure
BC> S> set it back after you're done though!

The user's screen ht is an important enough (and commonly used enough)
variable to warrant having its own 0-99 type fixed 'getuser'.

I was surprised to see WIDTH got one, but HEIGHT got ignored.

Also, we would NEVER have to worry about 1100468 changing in some future
release, breaking the doors that used it.
 
--1445--
      
BB> Please UPLOAD all packets which you say GURU TOSS.  I can not work
BB> BLINDLY.  A 100K stack is NOT the answer.  I'm sure TOSS uses less than
BB> 10K of stack in actuality.  The only thing a large stack gets you is more
BB> SLOP memory.

I've been keeping a log of stack use with "StackCheck V1.0 by Gunther
Rohrich".

CNetToss Wed 26-Jan-94 05:14:50   Used max:  5570
CNetToss Wed 26-Jan-94 15:42:31   Used max:  5574
CNetToss Wed 26-Jan-94 15:55:54   Used max:  5546
CNetToss Wed 26-Jan-94 15:57:07   Used max:  5570
CNetToss Wed 26-Jan-94 16:29:32   Used max:  5558
CNetToss Wed 26-Jan-94 21:56:37   Used max:  5570
CNetToss Thu 27-Jan-94 04:57:46   Used max:  5574
CNetToss Thu 27-Jan-94 07:37:32   Used max:  5558
CNetToss Thu 27-Jan-94 07:43:33   Used max:  5558
CNetToss Thu 27-Jan-94 14:56:25   Used max:  5570
CNetToss Thu 27-Jan-94 16:43:32   Used max:  5558
CNetToss Thu 27-Jan-94 16:44:39   Used max:  5558

That's about all the stack I EVER see CNetToss use.

Perhaps someone else will give it a try.  (I do NOT have 5-10 networks, 2meg
packets, 2000 user accts, 3000 subs, etc.)


Here's CNet's stack use....
CNet:BBS Port 0 Wed 26-Jan-94 17:38:40   Used max:  3636
CNet:BBS Port 0 Wed 26-Jan-94 17:40:21   Used max:  3636
CNet:BBS Port 0 Wed 26-Jan-94 17:44:21   Used max:  3636
CNet:BBS Port 0 Wed 26-Jan-94 17:46:13   Used max:  6672
CNet:BBS Port 0 Wed 26-Jan-94 17:48:10   Used max:  3636
CNet:BBS Port 0 Wed 26-Jan-94 17:50:11   Used max:  3636
CNet:BBS Port 0 Wed 26-Jan-94 17:51:01   Used max:  3636
CNet:BBS Port 0 Wed 26-Jan-94 17:56:34   Used max:  5440
CNet:BBS Port 1 Wed 26-Jan-94 18:00:41   Used max:  5658
CNet:BBS Port 0 Wed 26-Jan-94 18:01:47   Used max:  3636
CNet:BBS Port 1 Wed 26-Jan-94 18:06:59   Used max:  5658

Again, I can't get it to ever use more than 3-10K of stack.

Perhaps someone else will give it a try.  (I have not monitored 20 lines,
20000 files, CD-ROMs, giant yanks and QWK packets, etc.)
 
Remember, setting your stack TOO high (100-200K) can waste valuable
memory that you may need for that (or other) programs.  (Causing THEM
to have low-memory problems themselves.)

--1446--

I hope it is OK to mention this in public.  Ken mentioned it to beta-testers
and I don't know if its a top-secret thing or not.  (I hope not, nothing worse
than watching someone struggle with something that will shortly be a thing of
the past anyway.)

TrapWake won't be needed soon.  CNet itself will answer fidonet calls and call
trapdoor itself.

I don't know how soon this will happen, so I don't know if it's worth spending
too much time trying to set-up TrapWake, just to remove it all afterwards.
 
--1447--

On Mon 24-Jan-1994  8:46p, Pete Baker said to David M. Weeks:
PB> Ok, just not where I thought it would be.. and since it was not listed in
PB> the INDEX (hint hint)

On Sat 22-Jan-1994 10:38a, Mike Fitzgerald said to Pete Baker:
MF> Page 265. Could use  better index/table of contents.

On Sun 23-Jan-1994  8:24p, Dave Anderson said to Pete Baker:
DA> I seen it somewhere.  Dont remember where, since the Index is so
DA> incomplete.


If anyone is interested in helping me put together a 2nd "very
expanded" index for the CNet v3.0 manual, please respond here.

I'll need about 5 volunteers.

We'll split the manual's 285 pages onto 5 equal parts.

Each volunteer will be responsibly for creating just his section.
> Volunteer #1 : Pages   1-57
> Volunteer #2 : Pages  58-114
> Volunteer #3 : Pages 115-171
> Volunteer #4 : Pages 172-221
> Volunteer #5 : Pages 229-285

Just start on the first page of your page-group, and type-in any
key-words that you feel are significantly mentioned on that 1 page.

Send me a file like:
> 115 Feedback/call, Editor lines, Max email, inactivity-days...
> 116 Daily pfile-minutes, sending log, sysop-flag...
> 117 Send/rec email, mail-expiration, bulk-mail, party/multi-mail...
> ...
> 170
> 171

Each line starts with the page #.
Each keyword is separated with a comma.

Do NOT start on this project until AFTER we've decided who is doing which
page #s.

I'll gather, join, de-dup, and sort all 5 of index parts and make it
available to everyone.

I'll even write a cmd that'll search the index-file for you.
Typing "INDEX joinlink" (online or offline) will instantly tell you
which pages of the manual cover the topic of JOINLINKing.

--1448--

After switching from v3.01beta to v3.05, my SAM totals are:

>                      Last   Setup  Period   Total Current
> Feedback               0       0      12      32       0
> Mail sent              0       0    2893    3350     547
> Mail to #1             0       0    1805    2130      41
> Posts                  0     1.2G    1.6G    1.6G   7766
> Responses              0     692M    1.6G    1.6G     31K

Amaint, pointers, count, etc, do NOT seem to fix them.

Any ideas?

--1449--

I wish all things were this small...

Hitting TAB in CONFIG/EVENTS does NOT skip around to the various
string-gadgets IN ORDER.

--EOF--

Sunday 30-Jan-94 08:19:55

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

