Ticket History #213: document libc and gssrpc incompatability with regard to NIS

From hartmans@MIT.EDU Wed Nov 20 20:19:19 1996
Received: from MIT.EDU (SOUTH-STATION-ANNEX.MIT.EDU [18.72.1.2]) by rt-11.MIT.EDU (8.7.5/8.7.3) with SMTP id UAA28841 for <bugs@RT-11.MIT.EDU>; Wed, 20 Nov 1996 20:19:19 -0500
Received: from LOLA-GRANOLA.MIT.EDU by MIT.EDU with SMTP
id AA28169; Wed, 20 Nov 96 20:19:17 EST
Received: (from hartmans@localhost) by lola-granola.MIT.EDU (8.6.12/8.6.12) id UAA14086; Wed, 20 Nov 1996 20:19:16 -0500
Message-Id: <199611210119.UAA14086@lola-granola.MIT.EDU>
Date: Wed, 20 Nov 1996 20:19:16 -0500
From: Sam Hartman <hartmans@MIT.EDU>
Reply-To: hartmans@MIT.EDU
To: krb5-bugs@MIT.EDU
Cc: mycroft@MIT.EDU
Subject: document libc and gssrpc incompatability with regard to NIS
X-Send-Pr-Version: 3.99



System: NetBSD lola-granola 1.1B NetBSD 1.1B (LOLA) #2: Thu Jul 11 00:13:13 EDT 1996 mycroft@zygorthian-space-raiders:/afs/sipb.mit.edu/project/netbsd/dev/current-source/build/i386_nbsd1/sys/arch/i386/compile/LOLA i386


On platforms (excluding Athena Ultrix for some reason)

that have NIS in libc but do not have shared libraries, a conflict
exists that requires you to cause additional symbols to be resolved to
make sure all of
the RPC code comes from libgssrpc instead of libc for kadmin.

Mark and Tom have run into this; we should document
because we can't really fix.





From: Tom Yu <tlyu@MIT.EDU>
To: hartmans@MIT.EDU
Cc: krb5-bugs@MIT.EDU, mycroft@MIT.EDU, krb5-prs@RT-11.MIT.EDU
Subject: Re: krb5-doc/213: document libc and gssrpc incompatability with regard to NIS
Date: Sun, 24 Nov 1996 22:00:30 -0500

There has been some debate over whether this namespace collision will
cause us to hold up the release.

We should probably decide on this soon.

---Tom

From: "Theodore Y. Ts'o" <tytso@MIT.EDU>
To: Tom Yu <tlyu@MIT.EDU>
Cc: hartmans@MIT.EDU, krb5-bugs@MIT.EDU, mycroft@MIT.EDU,
krb5-prs@RT-11.MIT.EDU
Subject: Re: krb5-doc/213: document libc and gssrpc incompatability with regard to NIS
Date: Mon, 25 Nov 1996 10:44:21 -0500

Date: Sun, 24 Nov 1996 22:00:30 -0500
From: Tom Yu <tlyu@MIT.EDU>

There has been some debate over whether this namespace collision will
cause us to hold up the release.

We should probably decide on this soon.

Would anyone like to symmarize the case why it *should* cause us to hold
up the release?

The alternative is to document that if you have a static libc which has
NIS, you should configure --with-netlibs=resolv where libresolv.a is a
resolver library that doesn't have any of the NIS/YP crap in it....

- Ted

From: mycroft@gnu.ai.mit.edu (Charles M. Hannum)
To: "Theodore Y. Ts'o" <tytso@MIT.EDU>
Cc: Tom Yu <tlyu@MIT.EDU>, hartmans@MIT.EDU, krb5-bugs@MIT.EDU,
krb5-prs@RT-11.MIT.EDU
Subject: Re: krb5-doc/213: document libc and gssrpc incompatability with regard to NIS
Date: 25 Nov 1996 11:14:34 -0500

"Theodore Y. Ts'o" <tytso@MIT.EDU> writes:


Because it's bloody stupid to have to replace the resolver? krb5 is
practically its own OS already.


From: "Theodore Y. Ts'o" <tytso@MIT.EDU>
To: mycroft@gnu.ai.mit.edu
Cc: "Theodore Y. Ts'o" <tytso@MIT.EDU>, Tom Yu <tlyu@MIT.EDU>,
hartmans@MIT.EDU, krb5-bugs@MIT.EDU, krb5-prs@RT-11.MIT.EDU
Subject: Re: krb5-doc/213: document libc and gssrpc incompatability with regard to NIS
Date: Mon, 25 Nov 1996 11:39:03 -0500

We already have to replace the vendor OS resolver in some cases where
the resolver is buggy. All we would be doing here is extending the
definition of "buggy" to include "supports NIS/YP", which really isn't
all that far of a stretch. :-)

In fact, in a number of cases having a libc which supports NIS/YP means
that often means that the libc's resolver is screwing up gethostbyname()
anyway, so you end up having to replace the resolver in that case
already.

- Ted


From: Tom Yu <tlyu@MIT.EDU>
To: "Theodore Y. Ts'o" <tytso@MIT.EDU>
Cc: Tom Yu <tlyu@MIT.EDU>, hartmans@MIT.EDU, krb5-bugs@MIT.EDU,
mycroft@MIT.EDU, krb5-prs@RT-11.MIT.EDU
Subject: Re: krb5-doc/213: document libc and gssrpc incompatability with regard to NIS
Date: Mon, 25 Nov 1996 14:38:37 -0500



Do we know for certaint that --with-netlibs=resolv actually works? I
haven't had the time to try it yet.

---Tom

From: Doug Engert <deengert@anl.gov>
To: Tom Yu <tlyu@MIT.EDU>
Cc: "Theodore Y. Ts'o" <tytso@MIT.EDU>, hartmans@MIT.EDU, krb5-bugs@MIT.EDU,
mycroft@MIT.EDU, krb5-prs@RT-11.MIT.EDU
Subject: Re: krb5-doc/213: document libc and gssrpc incompatability with regard to NIS
Date: Mon, 25 Nov 1996 13:49:09 -0600

Tom Yu wrote:

I build beta 7 for the SunOS 4.1.x with a -with-netlibs=resolve2 which
is a modified resolver lib which goes directly to DNS. It works
with this system.

Doug Engert



From: Tom Yu <tlyu@MIT.EDU>
To: hartmans@MIT.EDU
Cc: krb5-bugs@MIT.EDU, mycroft@MIT.EDU, krb5-prs@RT-11.MIT.EDU
Subject: Re: krb5-doc/213: document libc and gssrpc incompatability with regard to NIS
Date: Mon, 25 Nov 1996 16:31:39 -0500

Mycroft claims (by way of Sam) that getservbyname() will still pull in
NIS. For some reason this didn't make it into the PR.

---Tom

State-Changed-From-To: open-closed
State-Changed-By: tlyu
State-Changed-When: Sun Feb 22 21:14:05 1998
State-Changed-Why:

This is moot now that gssrpc symbols have been mass-renamed.