On Wed, 2002-04-24 at 11:48, Igor Stroh wrote:
> On Wed, 2002-04-24 at 16:56, David E. Storey wrote:
> > attributetype ( 1.3.6.1.4.1.7914.1.2.1.14 NAME 'qmailAccountPurge'
> >         DESC 'The earliest date when a mailMessageStore will be purged'
> >         EQUALITY generalizedTimeMatch
> >     ORDERING generalizedTimeOrderingMatch
> >         SYNTAX 1.3.6.1.4.1.1466.115.121.1.24
> >         SINGLE-VALUE )
> 
> Afaik generalizedTimeMatch has a NO-USER-MODIFICATION flag, I'd like to
> set the timestamp manually though... also, the `date +'%s'` format is
> easier to compare then generalized time... just my two cents :)

Syntaxes only have DESC flags. createTimestamp and updateTimestamp have
the NO-USER-MODIFICATION flag associated with them at the attributeType
definition, but nowhere in the attributeType above did I specify that
flag.

You are correct that using unix epoch seconds is MUCH more efficient and
faster than using an LDAP generalizedTimeStamp. (And I knew this would
come up...) Unfortunately, it just isn't standard. It's like using
NUMBER in SQL versus a DATE. Sure it will be faster, but it would be
MUCH more useful to store the data in a form that is much more useful
for it. (I'm working with a database now that stores epoch seconds
rather than DATE and it is a real pain to work with it. I've since
mirrored the DB over and converted the columns and all is MUCH nicer
now.)

d!

Attachment: signature.asc
Description: This is a digitally signed message part

Reply via email to