Re: I-D ACTION:draft-ietf-ldup-lcup-04.txt

Kurt D. Zeilenga wrote:

>At 11:43 AM 3/10/2003, Rich Megginson wrote:
>  
>
>>It is definitely not the intention of the authors of LCUP that the server has to store any histories of content updates.
>>    
>>
>[...]
>  
>
>>A state based system can store historical state.  One of the assumptions of LCUP is that the server already has some sort of underlying historical information storage model, like a log or a state-based system with historical information.
>>    
>>
>
>How do you reconcile your statement that a server need not store
>any histories with your statement that LCUP assumes an historical
>information storage model?  They seem counter to each other.
>
Another misunderstanding.  I meant that we did not intend a server 
implementing LCUP to have to store any information specific to a client, 
like the last content update for a particular client.

>>>I disagree.  Every technical specifications limit implementation
>>>approaches to those which can produce conformant implementations.
>>>However, standard track technical specifications should avoid
>>>placing undue constraints on implementation details.
>>>      
>>>
>>Define "undue".
>>    
>>
>
>"inappropriate" or "improper". 
>
I just don't see how ANY synchronization protocol that attempts to not 
send every entry every sync session can be useful or even possible if 
the server stores no historical (even very recent) state information.

>  
>