On 7/30/2010 11:42 AM, Andrew Deason wrote:

> Which... seems to be the cell name, not realm. I'm assuming this
> originates from the time when cell == realm. For modern stuff we need to
> iterate through afs_krb_get_lrealm and use the longest one, right?
> 
> And "argh": what if the administrator configures another realm to be
> 'local' and it has a longer name? I presume there's nothing we can do
> about that... except perhaps documenting it?

My personal opinion is that it is not appropriate for cell A to second
guess what cell B may or may not accept.  I would remove this check
entirely.

Jeffrey Altman


Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

Reply via email to