Re: should guidelines say something about NOT using IMPLIED?

["Randy Presuhn" <randy_presuhn@mindspring.com>]

Hi -

> From: "Andy Bierman" <abierman@cisco.com>
> To: "C. M. Heard" <heard@pobox.com>
> Cc: "Mreview (E-mail)" <mreview@ops.ietf.org>
> Sent: Thursday, February 06, 2003 12:57 PM
> Subject: Re: should guidelines say something about NOT using IMPLIED?
...
> It is case-sensitive, but alphabetical.
...

It does *not* alphabetize correctly.  Even if one ignores the fact that
humans expect case-sensitive alphabetical sorts to put the "a" and
"A" words together, there are numerous other problems.  For example,
it won't put Schön *before* Schoepfung where it belongs.  It doesn't
even come close to producing alphabetical results for French or
German. The (expired) draft draft-faltstrom-i18n-sorting-*.txt
spelled out in gory detail why the kind of ordering SNMP GetNext
provides isn't alphabetical.

We wouldn't have as many MIB writers trying to use IMPLIED if we
could stamp out the myth that it was somehow useful.

Randy


Randy