[ 
https://issues.apache.org/jira/browse/COLLECTIONS-825?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Claude Warren updated COLLECTIONS-825:
--------------------------------------
    Description: 

{noformat}

src/main/java/org/apache/commons/collections4/bloomfilter/SimpleHasher.java
     * </p>
     * @return The default increment to use when the requested increment is 
zero.
     */
    public long getDefaultIncrement() {
Member
{noformat}

@aherbert aherbert on 27 Feb

I think this should be dropped. The default increment will fix the rare case 1 
in 2^64 that the increment is zero. But it does not fix the more common case (1 
in nbits) that the increment for the shape will be zero. Dropping this 
simplifies the constructor. A zero increment can be handled later.
@Claude-at-Instaclustr Claude-at-Instaclustr on 10 Mar

This is to handle the case where an existing implementation has chosen a 
different default value while still allowing the commons code to drop in place.


> BloomFilter: drop SimpleHasher.getDefaultIncrement 
> ---------------------------------------------------
>
>                 Key: COLLECTIONS-825
>                 URL: https://issues.apache.org/jira/browse/COLLECTIONS-825
>             Project: Commons Collections
>          Issue Type: Improvement
>            Reporter: Claude Warren
>            Priority: Major
>
> {noformat}
> src/main/java/org/apache/commons/collections4/bloomfilter/SimpleHasher.java
>      * </p>
>      * @return The default increment to use when the requested increment is 
> zero.
>      */
>     public long getDefaultIncrement() {
> Member
> {noformat}
> @aherbert aherbert on 27 Feb
> I think this should be dropped. The default increment will fix the rare case 
> 1 in 2^64 that the increment is zero. But it does not fix the more common 
> case (1 in nbits) that the increment for the shape will be zero. Dropping 
> this simplifies the constructor. A zero increment can be handled later.
> @Claude-at-Instaclustr Claude-at-Instaclustr on 10 Mar
> This is to handle the case where an existing implementation has chosen a 
> different default value while still allowing the commons code to drop in 
> place.



--
This message was sent by Atlassian Jira
(v8.20.7#820007)

Reply via email to