Re: ;binary transfer of the binary syntax

[hahnt@us.ibm.com]

Kurt,

My preference is that ";binary" is NOT required to be sent (nor should it be expected).  I would prefer to see "language" along the lines of:

whatever is sent in on the request is what will be returned (like attribute type alias names).  If the specific attribute type name is not specified, then the "no description" form is returned.  Thus, if

foo

is requested,  the response will contain

foo

If

foo;binary

is requested, the response will contain

foo;binary

If no attributes (i.e. all attributes), or "*" is requested, the response will contain

foo

(among others).  If

fooAliasName

is requested, the response will contain

fooAliasName

If

fooAliasName;binary

is requested, the response will contain

fooAliasName;binary

Regards,
Tim Hahn

Internet: hahnt@us.ibm.com
Internal: Timothy Hahn/Endicott/IBM@IBMUS or IBMUSM00(HAHNT)
phone: 607.752.6388     tie-line: 8/852.6388
fax: 607.752.3681

Sent by:        owner-ietf-ldapbis@OpenLDAP.org

To:        ietf-ldapbis@OpenLDAP.org
cc:        
Subject:        ;binary transfer of the binary syntax



Question:  Should an attribute of syntax binary be transferred
with or without the ;binary option?

That is, if I have
       ( 1.2.3 NAME 'foo' SYNTAX 1.3.6.1.4.1.1466.115.121.1.5 )

Should foo be transferred as "foo" or as "foo;binary"?

It seems to me one or the other should be mandated.  I would
prefer that just "foo" be used as "foo;binary" is redundant.
However, inetOrgPerson already has attributes of this syntax
and requires ;binary transfer of them.

I've noticed other schema items (such as Netscape's nsLIdata)
is published as syntax binary, but applications using this
attribute don't support or expect ";binary".

Seems we have a bit of an interoperability problem if we
don't specify consistent transfer defaults for LDAP syntaxes.

Is this something we need to address?  If so, how?

Kurt