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

Re: Duplicate entries filter

At 05:44 PM 8/15/00 +0100, David Chadwick wrote:
>i.e. it was to act on the way that the results of the search operation 
>were returned, but was not meant to affect the operation of the 
>search filter. If one considers that in many cases many thousands 
>of entries are typically filtered down to only a handful, then 
>duplication of the results with this control is only a small task, as 
>opposed to a) where we have duplication of the thousands of 
>entries and then the application of the filter to each of these. 

a) sounds like lots more work, but in reality may not be (due to
indexing and/or other optimizations).  The server uses indexing
(as normal) to locate candidates, duplicates the candidates,
applies the filter, returns the results.

However, this shouldn't be the primary issue of a) v. b) (or both).

The issue should be what behavior do applications need and whether
a) or b) (or both) or other controls fulfill the need.  I think
we should take a step back and discuss background and intended
use information before choosing.

Kurt