T.R | Title | User | Personal Name | Date | Lines |
---|
93.1 | | HARE::STAN | | Mon Feb 11 1985 12:53 | 5 |
| If you sent this message from a REMACP.COM file, then that is why
you're getting the protocol not supported message.
The guy talking to you is talking REMACP protocol and you're
sending him back straight ASCII. You'll either have to talk
REMACP protocol or find some other way to send this message.
|
93.2 | | HARE::STAN | | Mon Feb 11 1985 12:55 | 2 |
| How about a check in the system login file - if the connection
is remote and interactive, print a message and then LOGOUT.
|
93.3 | | SNOV10::QUODLING | | Mon Feb 11 1985 20:55 | 3 |
| Why doesn't he get more micom ports, or better still a LAT!
Q
|
93.4 | | KOALA::ROBINS | | Tue Feb 12 1985 12:31 | 4 |
| re .2: that can be bypassed with "Username/co=not_the_system_login_file"
Scott.
|
93.5 | | FDCV07::NORRIS | | Tue Feb 12 1985 12:46 | 6 |
|
Re -1.
Set up the account captive and defcli.
Ed
|
93.6 | | EAYV04::PETERM | | Sat Feb 16 1985 17:16 | 7 |
| In a similar vein, I have no problem with people setting host to
other machines on this site. But in allowing SET HOST <Ayr machine>
I implicity allow SET HOST <New England> which I definitely do not
want my users to do. Is there any way to limit this. Even being
able to specify this area only would be an improvement.
PeterM
|
93.7 | | FKPK::KONING | | Mon Feb 18 1985 20:34 | 5 |
| If you don't put the nodenames into your database, you'd make it a lot
harder (but not impossible). You could also define n000 logical names
for all the "forbidden" nodes pointing off to something harmless.
Paul
|
93.8 | | NUHAVN::CANTOR | | Sun Feb 24 1985 01:45 | 6 |
| re .4, .5
If SYS$SYLOGIN is defined, a login attempt with /NOCOMMAND will NOT bypass
the system-wide login file, whether or not the account is captive.
Dave C.
|
93.9 | | JON::MORONEY | | Fri Mar 08 1985 10:31 | 5 |
| What is the advantage of doing a set host 0 on a slow machine? Wouldn't
the overhead of REMACP make things (slightly) worse?
-Mike
/\@@/\
|
93.10 | | FKPK::KONING | | Fri Mar 08 1985 12:09 | 7 |
| If you use Set Host, your QIOWs to SYS$OUTPUT aren't made to wait until
the last byte has been given to the terminal Mux, as is the case in the
non-network mode. VMS unfortunately copied that design from RSX, and has
been suffering from it evern since, or at least until V3. Did that finally
get fixed in V4? I've heard rumors...
Paul
|
93.11 | | SPRITE::OSMAN | | Tue Apr 23 1985 15:44 | 20 |
| The very first message in this note confuses me. The writer says users are
saying
SET HOST 0
to avoid the crowded MICOM.
I don't understand. Perhaps the writer meant to say users are saying
SET HOST X
from some other system "Y" ?
It seems to me that "SET HOST 0" can't possibly help, since you're already
logged in, and hence why bother SETting host at all ? Just get to work!
Confused,
/Eric
|
93.12 | | VIKING::WASSER_1 | | Tue Apr 23 1985 17:45 | 10 |
| re. .11: I think the person recommending "SET HOST 0" may have
desired to use several accounts on the same machine. This
method gives you owner access to your other account without
forcing you to log off and face a crouded teminal switch. I
don't know if "SET HOST 0" is better than "SET HOST node" where
"node" is your current node name. Is there a special case
for "0" where the login does not have to go through DECnet?
-John A. Wasser
|
93.13 | | PARVAX::PFAU | | Tue Apr 23 1985 21:47 | 4 |
| 0 just means the current node. If I'm on PARVAX, there's no difference
whether I type SET HOST 0 or SET HOST PARVAX.
tom p
|