Re: Problems with EQUALITY generalizedTimeMatch

[Jon Roberts <man@mentata.com>]

Quanah Gibson-Mount wrote:
> --On Tuesday, February 17, 2004 12:41 PM -0600 Jon Roberts 
> <man@mentata.com> wrote:
> > Try an ORDERING clause to your attribute definitions, like:
> >
> > ORDERING numericStringOrderingMatch
> >
> > See also:
> >
> > http://www.openldap.org/doc/admin22/schema.html#Extending%20Schema
> >
> > which I notice doesn't list the generalizedTimeMatch syntax.
>
>
>        ORDERING generalizedTimeOrderingMatch
>
> is the syntax.

Right. Also not listed in the Admin Guide (or in this case even found by 
a search of the list archives until perhaps this thread), although 
plainly in RFC 2252 along with other syntaxes. The numericStringOrdering 
wouldn't work with time codes, I assume, because of the suffix. Once 
again, my mistake; I glanced when I should've looked.

Incidentally, I've been able to effectively sort and subrange UTC 
signatures with ORDERING caseIgnoreOrderingMatch. Since an advantage of 
UTC is a natural numeric sort, can anyone briefly explain the efficiency 
advantages of the generalizedTime syntaxes?

Jon Roberts
www.mentata.com