I have an improved klog now:
/afs/umich.edu/group/itd/build/mdw/tmp/openafs-143-klog2.diff

Over the last version this has:

/1/ this version actually built & tested with 1.4.3fc3
        a recent ancestor actually worked with heimdal.

        Note: source says "USING_HEIMDAL" which only exists in the rxk5
        patch.  For vanilla builds, you'll need to choose one of these:
        HAVE_KRB5_CREDS_SESSION
        HAVE_KRB5_PRINCIPAL_GET_COMP_STRING
        HAVE_KERBEROSV_HEIM_ERR_H
        or invent a configure test for absence of krb5_get_prompt_types().
        I recommend "HAVE_KRB5_CREDS_SESSION" if you're not sure which.
        The configure test would be most in keeping with the prevailing
        code style.

/2/ add_etbl.c - is a chunk extracted out of aklog_main.c.
        This should probably be shared with aklog.
        ( note#2a: this is part of the general libcom_err mess. )
        You may need a different solution with heimdal,
        but probably not.

/3/ Makefile.in
        patch more completely fleshed out.
        - libraries completely reordered.  I think I know who
        wrote this, and he's a very nice guy and a wonderful host.
        But his library ordering is, uh, confused.  Mine might
        be too, but it works with klog and should with aklog
        ( sadly my aklog build broke - see note#2a )
        - install rules.  Since I didn't do anything to break
        kaserver, this makefile will just overwrite the klog installed
        by kauth.

/4/ -524 is now -unwrap, works, and has more evil.

/5/ vice id lookup actually works right now.

/6/ skipwrap - uses more unsigned math to avoid unexpected wraparound.

                                -Marcus
_______________________________________________
OpenAFS-devel mailing list
[email protected]
https://lists.openafs.org/mailman/listinfo/openafs-devel

Reply via email to