RE: schema-07 comments

["Kathy Dally" <kdally@mitre.org>]

Hi Hallvard!

As before, see "KLD2:".

Thanks,
Kathy Dally

-----Original Message-----
From: Hallvard B Furuseth
[mailto:h.b.furuseth@usit.uio.no] 
Sent: Wednesday, June 02, 2004 5:26 PM
To: Kathy Dally
Cc: ietf-ldapbis@OpenLDAP.org
Subject: RE: schema-07 comments


Kathy Dally writes:
>Hallvard B Furuseth:

>> 2.23  postalAddress
>> 2.27  registeredAddress
> 
>> "15 Main St., Ottawa, Canada"
> 
> Please use the LDAP syntax, "15 Main
> St.$Ottawa$Canada".
> kld:  No.  The example is a single address.

Eh?  "$" doesn't separate different addresses, it
is a separator
between lines of one address.  Look at the Postal
Address examples
in [Syntaxes].

> Escaping the ","s is fixed.

What escaping?
KLD2:  Oops!  I wonder what I was looking at?
Fixed.

>> 2.43  x500UniqueIdentifier
> 
>>   In X.520 [X.520], this attribute type is
called 
>>   uniqueIdentifier.  This is a different
attribute type from both the 
>>   "uid" and "uniqueIdentifier" attribute types.
>                ^^^^^^^^^^^^^^^^
> If you mean an LDAP "uniqueIdentifier" attribute
type, that is not
> defined in this document.  Where is it defined?
>
> kld: It is in RFC 1274.  However, we are trying
to avoid references to
> that RFC.  If "(uniqueIdentifier is specified in
RFC 1274.)", was
> added to the paragraph would that be ok?  Would
RFC 1274 have to be
> included as an Informative Reference?

I have no idea.  How about:

  This is a different attribute type from both the
"uid" and
  the obsolete "uniqueIdentifier" LDAP attribute
types.
      ^^^^^^^^                    ^^^^

> Also, see 2.39 in the I-D.

OK:

>   The uid attribute type contains computer
system login names
>   associated with the object.  (Source: RFC
1274,
>   RFC 2798).  Each name is one value of this
multi-valued attribute.

Hm.  I wouldn't call RFC 2798 (inetOrgPerson) a
source for uid, since
it's years newer than the use of uid in RFC 2253.
If you count RFCs
newer than uid as sources, isn't [Schema] itself
just as good a source?
KLD2:  Yes, in a sense.  That's why uid is
included in [Schema].

As far as I can tell, the source for uid is buried
in RFC 2253 which
says that LDAP 'uid' = X.500 'userid', combined
with RFC 1274 for
'userid'.

KLD2:  The name 'uid' is introduced in RFC 2253
(and its successor -DN-xx), but it does not
specify the attribute type.  The 'userId'
attribute type is in RFC 1274 and does not exist
in X.500.  A definition of 'uid' in a more recent
specification style is given in RFC 2798.
[Schema]
replaces the informative definition of RFC 2798,
as stated in section 1.1.


>> 1.1  Situation
>
>>   Section 3.4 of 
>>   this document supercedes the technical
specification for the 'dc' 
> kld:  not in my original

Fixed after submission, then.  I've just verified
that this sentence is
in the official -07 draft.
KLD2:  Huh?  I thought you were referring to a
formatting problem.  I see that the section should
be "2.4".  Is there anything else?


>> 7.1  Normative
> 
>> ...[ROADMAP]  Zeilenga, K., "LDAP:  Technical
>                Specification Road Map",
> 
> Kill the '...'.
> kld:  I don't understand what's wrong.

The '...' in front of '[ROADMAP]' should be '   ',
unless '...' has some
special meaning I don't know about.
KLD2:  Ahh!  I'm still learning the difference
between space and period!  Fixed.

>> Appendix A  Changes RFC 2256
> 
> s/Changes/Changes made since/.
> 
> There are more changes:
> 
> - Removed '{number}' (minimum lower bound?)
after the SYNTAX oid for
>   all attributes that had that.

Whoops, I think I should have said 'minimum
_upper_ bound'.
KLD2:  Yes, caught it.

-- 
Hallvard