Re: RE: PCLS classes deprecated in PCELS

I think that there really are two separate questions.
1) Should PCLS be completed as a standards track RFC?  As far as I know, 
everyone agrees that the answer to that is yes.
2) How should PCIMe be mapped to LDAP?  My concern is taht you and others 
are holding that mapping up to a standard of PCLS compatibility that PCIMe 
itself does not meet.  I may be mistaken, but I believe that the 
differences between PCIMe and PCIM are such that it is at best extremely 
difficult (and quite possibly impossible) to accurately map PCIMe to LDAP 
in a fashion compatible with PCLS.  My feeling is that concerns about PCELS 
should be in the context of support for PCIMe, not in the context of 
support for PCIM without PCIMe.

I can understand that you (and possibly others) disagreed with the rough 
consensus on PCIMe.  But it is unfair to the PCELS authors to base 
objections to their mapping on objections to the differences and 
incompatibilities between PCIM and PCIMe.

Yours,
Joel M. Halpern

PS: In contrast to my concern above, if one can accurately map PCIMe to 
LDAP in a way that is compatible with PCLS, then that would clearly seem 
desirable.

At 10:47 AM 9/12/2003 -0500, Larry S. Bartz wrote:
>Joel M. Halpern wrote, On 09/12/03 09:43:
>>Given the discussion about the relationship between PCLS and the proposed 
>>PCELS, allow me to ask a naive question please.
>>There were some significant structural changes introduced between PCIM 
>>and PCIMe.  They were introduced to address problems with the PCIM 
>>structuring when other folks tried to use that as a basis for work.
>>Folks seem (as I read the exchanges) to be objecting to the corresponding 
>>changes in the LDAP schema.
>>Are the folks who implemented PCLS and are concerned about the changes 
>>expecting to support PCIMe?
>>If you are expecting to support PCIMe, how would you expect to deal with 
>>the fact that there are structural changes in the underlying model?
>
>
>I disagreed (in this forum) with some of the strategies of PCIMe.
>My concerns were not accomodated in the discussion or the design.
>Ultimately, I made a personal choice not to pursue my points further,
>as I recognize that intelligent, experienced, and reasonable people
>may disagree without causing harm.
>
>As far as I was concerned, PCIM and PCLS were already accomplished
>work, work with which I agreed, and work upon which I could (and did)
>build further. I have not experienced "problems with the PCIM
>structuring", quite the opposite. I have found PCIM and PCLS to be
>elegant, thorough, and highly serviceable designs. My work does not
>require the restructuring and additional complexity of PCIMe. Neither
>does my work require the QoS-specific components of PCIMe and PCELS.
>
>I don't have a requirement for PCIMe. I don't anticipate modifying
>my work to accomodate PCIMe.
>
>I see PCIMe as a departure from PCIM's simplicity and elegance. I
>see PCIMe as one of many possible elaborations upon PCIM's model.
>Even with that unfortunate "Update" word in RFC 3460, I still consider
>PCIM (RFC 3060) to be an authoritative document which describes a
>viable model.
>
>In the same spirit, I really want to see PCLS become an RFC. And
>I especially don't want PCELS to clobber it or impinge upon it
>in any way.
>
>
>--
>--
>#::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|
># Larry Bartz
>#
>#  voice (317) 226-7060
>#  FAX   (317) 226-6378
>#::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|