"Daniel Richard G." <email address hidden> writes:
> Russ, there was no Launchpad weirdness going on; I filed the bug against
> kerberos-configs, since I thought that's where Kerberos-config-related
> bugs were supposed to go. (E.g. bug #369575 was moved there.) But of
> course, it concerns a file in the libpam-krb5 package.
Oh, okay. kerberos-configs is actually a regular source package that
provides the package krb5-config. It's only responsible for
/etc/krb5.conf and the logic to generate it, not for other configuration.
> What was the fix, then? Was it a code issue, or should the config have
> used one of the other options?
It wa a code issue. pam-krb5 had an incorrect implementation of
use_authtok due to my misunderstanding of the Linux PAM documentation,
which was fixed in 4.0 by changing the meaning of several options to more
closely match the intended behavior.
"Daniel Richard G." <email address hidden> writes:
> Russ, there was no Launchpad weirdness going on; I filed the bug against config- related
> kerberos-configs, since I thought that's where Kerberos-
> bugs were supposed to go. (E.g. bug #369575 was moved there.) But of
> course, it concerns a file in the libpam-krb5 package.
Oh, okay. kerberos-configs is actually a regular source package that
provides the package krb5-config. It's only responsible for
/etc/krb5.conf and the logic to generate it, not for other configuration.
> What was the fix, then? Was it a code issue, or should the config have
> used one of the other options?
It wa a code issue. pam-krb5 had an incorrect implementation of
use_authtok due to my misunderstanding of the Linux PAM documentation,
which was fixed in 4.0 by changing the meaning of several options to more
closely match the intended behavior.
-- www.eyrie. org/~eagle/>
Russ Allbery (<email address hidden>) <http://