[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: protocol-22 comments

At 03:40 AM 3/13/2004, Hallvard B Furuseth wrote:
>> The problem we current face is that effect of ordering is
>> generally unspecified.
>Hm?  I thought you just agreed that it _is_ specified.

The problem we currently face in the existing TS (RFC3377) is that
the effect of ordering is unspecified.

>I just noticed an error in 4.1.10 (Referral):
>>   The referral result code indicates that the contacted server does not 
>>   hold the target entry of the request.
>So does noSuchObject.

The quoted text is problematic in that a server which does hold
a target entry can return a referral.

>I suggest:
>    A referral indicates that the contacted server cannot perform the
>    operation, but one or more other servers can.
>I tried at first to formulate it in terms of naming contexts, but that
>got too clumsy.  Some operations cannot be performed if the server
>doesn't hold the naming context for the entry which the operation
>applies to, while others cannot be performed if it doesn't _master_ that
>naming context.

Right, referrals are be returned when the server chooses to
refer the operation to another server.