Path: utzoo!attcan!uunet!samsung!sdd.hp.com!zaphod.mps.ohio-state.edu!uwm.edu!csd4.csd.uwm.edu!adk
From: adk@csd4.csd.uwm.edu (Andrew D Kailhofer)
Newsgroups: comp.sys.ncr
Subject: Re: Tower 32 030001 & /etc/shadow
Summary: NCR V.3 level
Keywords: why does it not have one ?
Message-ID: <6890@uwm.edu>
Date: 11 Oct 90 04:37:23 GMT
References: <1109@cnw01.storesys.coles.oz.au>
Sender: news@uwm.edu
Reply-To: adk@csd4.csd.uwm.edu (Andrew D Kailhofer)
Organization: University of Wisconsin-Milwaukee
Lines: 19


The release definition (I think) says that this isn't there.  I know
that I've complained before about what NCR chooses to support and not
support, but the V.3 is *very close* to SVR3.2, with the exception of
a few little things, like the help thing, and like the /etc/shadow
thing.  In fact, not having the getspwent() stuff from libsec.a is no
great loss.  I can't think of any way to cause more incompatiblity
with old code than the way that AT&T implemented shadow passwords.
IMHO they should have modified the getpwent() set to "Do The Right
Thing" if /etc/shadow was there, and the previously right thing if it
wasn't.

I can only hope that somebody figures out the right way to do this.
Does anyone know what SVID-2 says?

Andy Kailhofer
Ameritech Services/Information Technologies
414/678-7793
a07932@gus.ameritech.com