Thanks to
Davis_at_Tessco.Com {
This is very likely just an oversight in code update. The utmp structure
got updated when the RFC's for hostnames changed, but not all of the code
that uses utmp got updated. If it's that important to you, you can request
through Compaq Services that it be fixed. This type of update is usually
not done immediately, but it's not released as a patch, it's included in
the next major release.
Alan Davis
}
Sure its important. If it wasn't why is it there?
magill_at_isc.upenn.edu {
This is an old bug. There is no trivial fix.
I don't know if it is finally addressed in 5.0.
}
original msg {
We have an alpha 3000/m400 with Digital Unix 40.e running on it and the
patch kit 01 installed.
i am running finger xxxx and that is what i am getting:
Login name: xxxx In real life: Fred Flintone
Directory: /users/cartoons/xxxx Shell: /bin/csh
Last login Tue Apr 27 18:46 on ttyp1 from ppp34-cartons.ya'
No Plan.
i can not see the hole fully qualified name of the host that opens the
connection. When a FQN does not exists and only the ip address is shown
there is no problem.
So i am looking for the fixed length of the array of chars in lastlog.h
because finger reads /var/adm/lastlog:
#define LL_LINELEN 8 /* terminal path length */
#define LL_HOSTLEN 16 /* hostname length */
struct lastlog {
time_t ll_time;
char ll_line[LL_LINELEN];
char ll_host[LL_HOSTLEN]; /* same as in utmp */
};
So the length is 16 and it can not represent the full name (why?). But
the frace "/*same as in utmp */" above and the contents of utmp.h makes
me wonder.
*utmp.h*
struct utmp {
....
....
char ut_host[64]; /* host name same as MAXHOSTNAMELEN */
};
}
take care
salaGatoR.
Received on Thu May 06 1999 - 07:16:20 NZST