[1488] in Kerberos-V5-bugs

home help back first fref pref prev next nref lref last post

Re: Kerberos V5 beta5 / DCE interoperability problem

daemon@ATHENA.MIT.EDU (Theodore Ts'o)
Mon Jun 19 15:44:40 1995

Date: Mon, 19 Jun 1995 15:43:00 +0500
From: Theodore Ts'o <tytso@MIT.EDU>
To: pato@apollo.hp.com
Cc: Theodore Ts'o <tytso@MIT.EDU>, KRB5-BUGS@MIT.EDU
Cc: "Doug Engert" <DEEngert@anl.gov>, sommerfeld@apollo.hp.com
In-Reply-To: Joseph N. Pato's message of Mon, 19 Jun 1995 10:38:36 -0400,
	<v02110102ac092057e3f1@[15.4.25.130]>

   Date: Mon, 19 Jun 1995 10:38:36 -0400
   From: pato@apollo.hp.com (Joseph N. Pato)

   >The correct convention is that the cell name should be upper case.  The
   >fact that DCE enforces the cell name to be in lower case, and then
   >forces the Kerberos realm name to match the cell name is a bug, which I
   >think they might try to fix in DCE 1.2.

   The cell name can be either upper case or lower case. I do not expect this
   to change any time in the future. We support a variety of global naming
   services - each with their own peculiarities about the way a name can be
   expressed. There is no guarantee that upper-case-only names are legal.

Are you sure about this?  I was assured that DCE forcibly lowercases the
cellname.  In fact, as I recall someone from HP (I'm pretty sure it was
you) protested approximately 6 months ago when Cliff proposed making an
Errata to the RFC which would force the realm names to be uppercase
(everything else would then be nonstandard).  The reason that was given
for our not doing this was because DCE wouldn't support lower case realm
names at all, no matter how hard an administrator tried to make it be
all upper case.

Doug, Bill, would you like to comment on whether this is true?

In any case, the deployed base uses all uppercase, and if it is true
that DCE can use all upper case realm names, I will recommend to Cliff
that we in the upcoming revision to RFC 1510, that we specify that the
realm name MUST be in all upper case.

						- Ted

home help back first fref pref prev next nref lref last post