uucico hang between sVr3.2 and sVr4

[Leslie Mikesell]

In article <1658 at svin02.info.win.tue.nl> debra at svin02.info.win.tue.nl (Paul de Bra) writes:

>When the sVr4 system calls the sVr3.2 system with uucico, everything
>goes well.
>When the sVr3.2 system calls the sVr4 system with uucico it hangs
>right after getting the Shere=thename^@ from the sVr4 system (where
>'thename' stands for the hostname on the sVr4 system).
>The uucico on the sVr3.2 system times out and exits.
>The uucico on the sVr4 system does not time out but just hangs
>indefinetely.

One possibility that is not entirely obvious is a syntax error in
the Permissions file (perhaps a continuation line where the previous
line does not have a \).  The reason it could act differently when
calling vs. being called is that the Permissions entry is located
by the MACHINE= entry on outbound calls, but on inbound calls it
is the LOGNAME= entry (and the first match is used).  For instance
if your first entry on the SysVr3.2 machine has LOGNAME=nuucp
and that is what the other machine logs in as, it wouldn't read any
farther in Permissions.  Then if the MACHINE= entry for the SysVr4
system is the last entry in the file and the syntax error is there
or somewhere in between, you get the silent death syndrome.  The reverse
could apply to the SysVr4 machine if they haven't fixed uucico to
give a reasonable error message.

Les Mikesell
  les at chinet.chi.il.us