Ticket History #127: v4kadmin testing assumes athena environment

From epeisach@MIT.EDU Tue Oct 22 08:10:21 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 IAA11879 for <bugs@RT-11.MIT.EDU>; Tue, 22 Oct 1996 08:10:20 -0400
Received: from KANGAROO.MIT.EDU by MIT.EDU with SMTP
id AA17831; Tue, 22 Oct 96 08:10:20 EDT
Received: by kangaroo.mit.edu; (5.65/1.1.8.2/08Mar96-0212PM)
id AA07539; Tue, 22 Oct 1996 08:10:19 -0400
Message-Id: <9610221210.AA07539@kangaroo.mit.edu>
Date: Tue, 22 Oct 1996 08:10:19 -0400
From: epeisach@MIT.EDU
Reply-To: epeisach@MIT.EDU
To: krb5-bugs@MIT.EDU
Subject: v4kadmin testing assumes athena environment
X-Send-Pr-Version: 3.99

mit

System: OSF1 kangaroo.mit.edu V3.2 214 alpha
Machine: alpha
The v4kadmind test suite right now assumes that
/usr/athena/bin/kpasswd exists. This is a bad assumption.

Look in config/unix.exp under the kadmin/v4server/unit-test
tree.
Either we should detect it - or we should compile our own local
version. I think enough of the source is present in the tree that
compiling a v4kadmin should be trivial....
(Note: Barry may have ideas on this already - then this is therefore a
marker indicating something we need to do)




From: "Barry Jaspan" <bjaspan@MIT.EDU>
To: epeisach@MIT.EDU
Cc: krb5-bugs@MIT.EDU
Subject: Re: krb5-admin/127: v4kadmin testing assumes athena environment
Date: Tue, 22 Oct 1996 18:07:44 -0400

Actually, I hadn't thought of this problem. You're right, though, the
kadmind4 unit tests do require a V4 kpasswd.

I'd prefer not adding more V4 kadmin sources to the tree if possible.
It is already very confusing to have a kadmin, kadmin.new, kadmin.v4,
and kadmin.old at the top level, along with lib/kadm and lib/kadm5,
and kadmin/server, kadmin/v4server, kadmin/v5server, kadmin/v5client,
and kadmin/v5passwdd. I don't even know what all these things are...

A site that cares about the kadmind4 unit tests will have a V4 passwd
around. So I think this should be found via configure, not built
locally, and not tested if it cannot be found or specified with a
--with option.

BTW, I think the --with options that control whether unit tests get
run should be mentioned in the error message from the Makefile when
they are not run. ie:

check-- check-runtest- check--tcl::
@echo "+++ Either tcl or runtest is unavailable. Kadm5 unit tests not run."
@echo "+++ Re-run configure with the --with-tcl or --with-runtest option to correct."

Barry


From: "Barry Jaspan" <bjaspan@MIT.EDU>
To: epeisach@MIT.EDU
Cc: krb5-bugs@MIT.EDU
Subject: Re: krb5-admin/127: v4kadmin testing assumes athena environment
Date: Mon, 4 Nov 1996 16:42:40 -0500

We decided to make this a high priority bug for 1.0 and resolve it by
making the kadmind4 tests not run by default. It turns out that was
already the case, though, because I left the DO_SUBDIRS rule out of
configure.in. Therefore, the high-priority-ness of this bug is fixed.

It would still be nice to make the unit tests search for kpasswd in an
intelligent manner or have kpasswd built locally, but that is not a
high priority for 1.0; Therefore, I've changed this PR to medium
priority.

Barry


From: Tom Yu <tlyu@MIT.EDU>
To: Barry Jaspan <bjaspan@MIT.EDU>
Cc: krb5-bugs@MIT.EDU
Subject: Re: krb5-admin/127: v4kadmin testing assumes athena environment
Date: Sun, 22 Feb 1998 21:13:31 -0500

`Tom Yu' made changes to this PR.

--- /tmp/gnatsa003n_ Sun Feb 22 21:13:06 1998
+++ /tmp/gnatsb003n_ Sun Feb 22 21:13:23 1998
@@ -16,7 +16,7 @@
->Severity: critical
+>Severity: serious

Responsible-Changed-From-To: bjaspan->krb5-unassigned
Responsible-Changed-By: mdh
Responsible-Changed-When: Tue Jun 23 23:04:31 1998
Responsible-Changed-Why:

bjaspan is no longer with the kerberos team