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

Re: draft-ietf-ldapext-ldapv3-vlv-02.txt last call



Jim Sermersheim wrote:

> I just noticed a couple things that should be changed in this doc.  Section 5 reads "...in the 
> supportedExtensions attribute", I think it's supposed to read "...in the supportedControl attribute".  

Changed. Thanks.

>And then in section 7, the sentence "The schema employed..." should either be removed or replaced with 
>something like what's in section 5.

Hmm. The subject is different. In section 5 we're talking about 
how clients tell whether or not a server supports the control.
How to tell this is defined in ldapv3. In section 7 we're
talking about how a client can determine whether a given
search it proposes to submit will be indexed or not. 
There's no prior art for how to do this. Netscape Directory
Server has a simple scheme and Netscape address book clients
use it to decide whether or not to turn on scrolling UI.
We intended to either publish that scheme, or work on 
a better one. The work to do either hasn't been done yet.
 
> Also (this is really picky), is there supposed to be a comma following the byOffset SEQUENCE in the request 
> control asn.1?

No.

There is a more serious problem with the VLV document, which
is that the "greaterThanOrEqual" relationship isn't the right
thing when the list is reverse sorted. I suspect that implementations
have done the "right thing" even though the document is wrong,
but clearly the document needs to be revised to reflect 
the correct behavior.