Re: Adding to the LDAP ACM to a WG charter

At 05:59 PM 2001-11-15, Chris Apple wrote:
>1) You are asserting that the IETF process won't
>  work for LDAP ACM and would need to be modified
>  slightly to result in the production of a good
>  LDAP ACM specification.

I see no need to change IETF process [RFC 2026] or
WG guidelines [RFC 2418] and am not advocating such.
The IETF process and guidelines are flexible.  The
process considers both working group and individual
efforts in producing RFCs.

Presently, I believe that LDAP ACM standardization
work is best pursued on an individual basis.

>2) You are advocating that this process modification
>  would be that multiple proposals for an LDAP ACM
>  should be written outside of the context of any WG.

I am suggesting that we let individuals pursue this work
under the established process.

>3) You are expecting one of these proposals to become
>  vetted in some way to become the eventual product
>  of a WG other than LDUP.

I am hoping that one or more proposals of appropriate quality,
drafted by individuals, would be submitted for consideration
for publication.  As with any other submission to the IETF,
the IETF would review these works and take appropriate action.

My assertions apply equally to LDUP as to any WG which might
consider taking on LDAP ACM work (or be formed specifically
for this work).  Any reference I made to LDUP was solely due
to the fact it happens to be the WG considering taking on
LDAP ACM work.  I don't think changing the WG venue matters
much (as the interested parties follow the work).

>4) You are asserting that the LDUP WG will either produce an
>  inferior or at least a delayed LDAP ACM than a different
>  WG with the same members would be able to.

I believe any WG taking on LDAP ACM work will be stalled due
to lack of consensus.  Individuals, not subject to WG
consensus, can produce an LDAP ACM for consideration by the
IETF much faster.

Of course, any LDAP ACM I-D submitted for consideration on
the Standard Track, whether produced by a WG or individuals,
has IETF consensus requirements.

And, in terms of meeting overall desires (powerful, flexible,
extensible, etc.), I think individuals can do better than
a WG in this case.  LDAPext had to remove functionality to
remove contention.  Individuals generally don't have to do
that.

Anybody remember the SGMP/CMOT wars which resulted in SNMP?
Basically I am suggesting a similar approach.  That is, let
each set of individuals develop the very best specification
they can, make "where's the ACM?" buttons ;-) and then
consider how to proceed.

Kurt

>>From: "Kurt D. Zeilenga" <[email protected]>
>>To: [email protected]
>>Subject: Adding to the LDAP ACM to a WG charter
>>Date: Thu, 15 Nov 2001 16:55:57 -0800
>>
>>
>>This message is intended to clarify the reasons why I
>>oppose having LDAP ACM work item on any IETF WG charter
>>at this time.
>>
>>Many (most?) of us (including I) would like very much
>>to see LDAP have a Standard Track powerful, flexible,
>>and extensible standard track access control model.
>>Many of us have tried very hard to produce such.  But
>>after years of trying, it became clear that the IETF
>>was not going to reach consensus on a number of critical
>>issues.  The issues are the same, the people are (mostly)
>>the same, a change of venue (LDAPext->LDUP) won't help.
>>
>>But a change to which process used may help.  Those who
>>support an LDAPext-ACM based solution should, as individuals,
>>produce the best damn LDAP ACM specification they can and
>>those who support alternative approaches should, as
>>individuals, produce the best damn specification they can.
>>We should give each set of individuals a reasonable amount
>>of time to produces these specifications.   Then, based upon
>>what is produced by the sets of individuals, determine how
>>best to proceed:
>>        a) adopt one (publish on Standard Track)
>>        b) adopt none
>>        c) create a WG to take one approach to Standard Track
>>
>>Based upon discussions I had at IETF#51 and with individuals
>>since then, I believe we will soon multiple pretty damn good
>>alternatives submitted (post IETF#52).
>>
>>IMO, we'll have an LDAP ACM must sooner than later if LDUP
>>stays out of way.
>>
>>Kurt
>>
>
>
>_________________________________________________________________
>Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp