Question: (I think I know the answer)
Is logging in from the outside via internet (eg thru msu gopher)
reserved for members, or can anybody do this?
Also, what systems (gophers, etc) is Grex "advertized" on?

30 responses total.

anyone can do this.

Currently, anyone who logs on to the msu-gopher can connect here.
It's just a matter of tracing through the menus.

As far as "advertising," I think the msu-gopher is the only system
that has a publicly accessible telnet connection here.  Of course
anyone can telnet here to cyberspace.org from any system if they
have access to the telnet prompt.  It's just that publicly accessible
systems usually limit telnet connections.

Someone slipped in ahead of me!

This response has been erased.

This response has been erased.

Anybody who has a modem and a phone line and about $18 a month to pay Netcom
for an account can telnet anywhere anybody else can telnet.  ~f ~p

Where is Netcom a local call from?

Send mail to info@netcom.com and their vacation program will send you
several screens' worth of stuff, including a list of phone numbers. 
They were originally just California, but they're in other parts of the U.S.
now also.  ISTR Texas, among others.  I don't know about Michican. 

Look in /u/rcurl/PDIAL.list for a description of netcom. Also of a hundred
or so other sites you can pay $$$ and telnet anywhere anybody can telnet.

login via the internet--that is what I'm doing, because I keep 
getting a busy signal when I dial in.  But w shows that only
ttyh0,1,2 and 4 are in use.  What happened to ttyh3 and 5 (I think there's 
a ttyh5 isn't there?  Inquiring minds want to know.

They are used. The dialins are relatively easy to access. (Don't know
just now which one I'm on now - probably not ttyh02, though - not noisy.

Yes there is and it's confusing. I am often the culptrit as I run a
program called layers which allows me to have multiple windows on Grex
all sharing the same phone line. It updates the temp file to show my sessions, 
but the line I'm connected on doesn't show up in the 'w' command.

You can get the same effect by running the unix screen program to have
multiple screens. the line really is in use though, and you can usually see
evidence of that by using 'finger' instead of 'w'. This will show you that each
of my sessions connected from ttyh3.

Oh, I didn't realize that finger worked like w I thought it was like f.  
Never assume with unix i guess.  Are you sure that is what's going on 
though Steve, w showed you (a bit later) as being on ttyh5 (and I think
also showed your other layers on ttyp(somethings)) but there was nobody
on ttyh3 the whole time (at least that w showed).

I don't know if this is the place for this, but I noticed a couple of newusers
idle quite a while: 

  3:29am  up 2 days, 13:58,  10 users,  load average: 2.94, 2.91, 2.52
User     tty       login@  idle   JCPU   PCPU  what
tdh      ttyh0     1:29am         1:06   1:06  /usr/local/games/lib/nh313adir/n
cicero   ttyh5     3:22am     7      5      1  /usr/local/bin/less -dre -P --Mo
jprice   ttyp1     2:27am    50      5         -sh 
newuser  ttyp2     2:07am  1:22                -
brianj   ttyp3     3:07am     6     25     13  beej 
bubbles  ttyp5     2:28am           52     16  w 
jeffrey  ttyp6     3:24am     1      5      4  trn 
chas     ttyp8     8:43pm         2:07   1:52  tf 
newuser  ttyp9     3:28am     1                -
newuser  ttypa     1:40am  1:46                -

tdh      ttyh0    Apr 18 01:29
cicero   ttyh5    Apr 18 03:22
jprice   ttyp1    Apr 18 02:27 (Z011000.PC3.AF.)
newuser  ttyp2    Apr 18 02:07 (unbvm1.csd.unb.)
brianj   ttyp3    Apr 18 03:07 (admiral.umsl.ed)
bubbles  ttyp5    Apr 18 02:28 (netcom6.netcom.)
jeffrey  ttyp6    Apr 18 03:24 (twain.ucs.umass)
chas     ttyp8    Apr 17 20:43 (NUBS63.ccs.itd.)
newuser  ttyp9    Apr 18 03:28 (cap.gwu.edu)
newuser  ttypa    Apr 18 01:40 (unbvm1.csd.unb.)

And they seem to be logged in by Internet. 

You mean that ttyh3 & ttyh5 are no longer in the hunt sequence?  Why not?

It has been observed before that if you dump your internet connection
without logging off grex, your connection ("process"?) stays around
until..what..staff dumps it? I presume that is what chas had. This is
very easy to do if you are surfing the internet, and opening connections
to hosts all over the place, until you have dozens piled up - and
then just dump your link. (How is the internet, or the internet hosts,
supposed to deal with this?)

I believe, in general, these days a dropped carrier will terminate a
connection MOST of the time.  I know I've done it before and been logged out.
At least it appeared I was logged out, because upon returning to grex a
few minutes later I wasn't still logged on from my previous
session.

Was that via the internet? My connection is a PPP link to the local
MichNet NAS, and grex has no way to know that I've dumped the PPP
link. I have found my process still active the day after I had
interrupted without logging off (which I try to avoid doing, but
can't do much about when the NAS quits on me!).

Ok, yes its through the internet via Merit to msu-gopher.
If *I* drop carrier it ends a session.  If, however, Merit or
msu-gopher 'locks up' and I have to drop carrier, then when I get back
on grex usually there *IS* an idle login sesion.

In theory, each place you're talking to along the way should notice that
you're not there & close its connection to the next one, allowing it to
do likewise.  Seems like it doesn't always work.

This response has been erased.

Its a case by case thing.  Sometimes I get logged off, others I'm still on.

Gopher locked up on me once when I was telnetted into Grex through that
freenet in Finland I used to go through before grex got added to the
gopher menu.  I logged in on a dialup a few hours later and noticed that
my connection from Finland was still there.  It must be somewhat inconsistent.

However it happens, are idle hung sessions enough of a bandwidth drain to
be worth writing a program to time out idle users? 

There's also an issue of cutting people off who are momentarily idle but
really still connected.  (If "momentary" is the word I want ...)  I have
had experience on another system: someone posted a question; I went into
the next room to check a reference book or two; when I came back to enter
the answer I was logged out.  I think the limit must have been under 5
minutes, and this was *really* irritating.  Better not go to the bathroom.

My terminal program has a little function that sends a CR or somesuch 
every few seconds to put the other computer "on hold" so to speak to prevent
just this very problem.  I've never used this feature though.  Boards that
I've been on that had an idle timeout also had a pause feature that would 
disable it.

As long as you don't do cicero's trick, the cost of maintaining an idle
session is nearly negligible. A few packets are passed to maintain the
connection (I'm told) but nothing of any consequence.

Has anyone considered that these migh be people who left their sessions
up when they abandoned newuser? I am sure this somnetimes happens.

re #13, I probalby logged off tyh3 and got back on ttyh5 later.
Yeah I'm positive that's what layers and screen do with the utmp file.

Re #15, No ttyh3 and h5 are still in the hunt group.
They are in use - as in tied up - even though they don't show
up in the utmp file. ttyuse knows that they are still in use, too.

Merit recently (around the end of May) required a unique nqme & password
to telnet out from the MSU connectio (and probably other connections as well)
Does anyone know of a way to get around this so people without Univ.
accounts can telnet to GREX (or wherever they wish to go !)

Given the large amount of people I still see on from burrow.cl.msu.edu, I
would doubt that authorization is needed to telnet out from there,
although it is needed to access Grex from the UM Gopher.  I'm not sure,
though; it's been a while since I've used Gopher since I got an MTS
account and that comes with authorization.

Absolutely. There are two ways to get to Grex via MSU-Gopher.
This simplest is by logging in as gopher and selecting menu items
13,5,7,8 in succession.
Alternately you can log in a web, and follow the hypertext to
Web Central at CERN, to WWW by subject, to subject "Computing", to Grex.
Use the "/" feature of lynx to search these long pages for what you
want. Once you access Grex's welcome page, there is a telnet link right there.

Response not possible - You must register and login before posting.

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss