[comp.unix.ultrix] DecServer 200/300 puts us into PASSALL mode sometimes

jdc@rama.UUCP (James D. Cronin) (11/08/90)

Followups to: comp.os.vms

We have a number of DecServer 200's and 300's used to connect to
a number of Vax/VMS, DecStation and Unix machines.  We use the
Ultrix LAT/Telnet gateway service to connect to Unix machines:
	"CONNECT TELNET NODE FRED DESTINATION SPOT"

For some unknown reason this puts us into passall mode more or
less at random.  Any ideas on why this happens?

Thanks...Jim Cronin
jdc@rama.sc.harris.com

D. Allen [CGL]) (11/08/90)

>We have a number of DecServer 200's and 300's used to connect to
>a number of Vax/VMS, DecStation and Unix machines.  We use the
>Ultrix LAT/Telnet gateway service to connect to Unix machines:
>	"CONNECT TELNET NODE FRED DESTINATION SPOT"
>For some unknown reason this puts us into passall mode more or
>less at random.  Any ideas on why this happens?

Any program on recent Ultrix that does "stty raw" sets the port to PASSALL.  
This includes the getty running at login time that reads your userid.

I asked on Oct 24 in comp.unix.ultrix how to turn it off; nobody knows.
-- 
-IAN! (Ian! D. Allen) idallen@watcgl.uwaterloo.ca idallen@watcgl.waterloo.edu
 [129.97.128.64]  Computer Graphics Lab/University of Waterloo/Ontario/Canada

grr@cbmvax.commodore.com (George Robbins) (11/09/90)

In article <1990Nov8.070438.6592@watcgl.waterloo.edu> idallen@watcgl.waterloo.edu (Ian! D. Allen [CGL]) writes:
> >We have a number of DecServer 200's and 300's used to connect to
> >a number of Vax/VMS, DecStation and Unix machines.  We use the
> >Ultrix LAT/Telnet gateway service to connect to Unix machines:
> >	"CONNECT TELNET NODE FRED DESTINATION SPOT"
> >For some unknown reason this puts us into passall mode more or
> >less at random.  Any ideas on why this happens?
> 
> Any program on recent Ultrix that does "stty raw" sets the port to PASSALL.  
> This includes the getty running at login time that reads your userid.
> 
> I asked on Oct 24 in comp.unix.ultrix how to turn it off; nobody knows.

In 99 out of 100 cases this is the correct behavior - to support unix
"raw" mode semantics, the server must **temporarily** enter passall
mode.  Since servers are not hard-wired terminal, there may be a few
cases where this doesn't give the intended result, but you break a
lot more stuff by not having it be the **default** behavior.

Please consider carefully before bitching about this, it's been broken
before (some 3.x releases) and took a long time to get it right again.

If the login behavior causes you distress, then what you really want
is /etc/ttys control of raw vs cbreak vs cooked mode for the prompt,
not LAT changes.

-- 
George Robbins - now working for,     uucp:   {uunet|pyramid|rutgers}!cbmvax!grr
but no way officially representing:   domain: grr@cbmvax.commodore.com
Commodore, Engineering Department     phone:  215-431-9349 (only by moonlite)

D. Allen [CGL]) (11/28/90)

>> Any program on recent Ultrix that does "stty raw" sets the port to PASSALL.  
>> This includes the getty running at login time that reads your userid.
>
>In 99 out of 100 cases this is the correct behavior - to support unix
>"raw" mode semantics, the server must **temporarily** enter passall
>mode.  Since servers are not hard-wired terminal, there may be a few
>cases where this doesn't give the intended result, but you break a
>lot more stuff by not having it be the **default** behavior.

I agree, but in that case don't we want PASSALL to *really* pass
everything?  As things stand, BREAK still gets you out.
-- 
-IAN! (Ian! D. Allen) idallen@watcgl.uwaterloo.ca idallen@watcgl.waterloo.edu
 [129.97.128.64]  Computer Graphics Lab/University of Waterloo/Ontario/Canada

pavlov@canisius.UUCP (Greg Pavlov) (11/29/90)

In article <1990Nov27.180637.8932@watcgl.waterloo.edu>, idallen@watcgl.waterloo.edu (Ian! D. Allen [CGL]) writes:
>>>Any program on recent Ultrix that does "stty raw" sets the port to PASSALL.  
      (etc...)
> I agree, but in that case don't we want PASSALL to *really* pass
> everything?  As things stand, BREAK still gets you out.

  ... yes, but some terminal servers including DEC's) allow you to disable
      and/or remap the BREAK trap  (I had to use this in a fun communications
      link that routed me thru three terminal servers scattered along two 
      leased lines and I needed to get the attention of/modify session paramet-
      ers in the 2nd and 3rd servers down the chain. Every time I used the link
      ......).

   greg pavlov, fstrf, amherst, ny
   pavlov@stewart.fstrf.org