0003380: Interval hash/Equal problem

Viewing Issue Simple Details [ Jump to Notes ]

[ View Advanced ] [ Issue History ] [ Print ]

Category

Severity

Reproducibility

Date Submitted

Last Update

0003380

[Squeak] Kernel

minor

always

03-29-06 01:28

05-05-06 22:10

Reporter

nicolas cellier

View Status

public

Assigned To

Priority

normal

Resolution

open

Status

new

Product Version

3.9

Summary

0003380: Interval hash/Equal problem

Description

The never ending story of hash equal problems:

(1 to: 3) = #(1 2 3). "is true"
(1 to: 3) hash = #(1 2 3) hash. "is false"

This does break Set behaviour:

self assert:
((Set new: 4) add: (2 to: 4 by: 2); add: #(2 4); size) =
((Set new: 5) add: (2 to: 4 by: 2); add: #(2 4); size)

Additional Information

Efficient Interval hash should fall back to slow super hash... why ? because they have same species as Array... and thus compare to true with Array.

Maybe all SequenceableCollection hash should be refactored, because they hash each and every element, which might be a slow on big collections.

There is a balance between fast hash code computation / slow Set access because of hash code collision.

Attached Files

Relationships

child of

0007002

new

The Mother of all issues involving interval

Notes
(0004884 - 29 - 41 - 41 - 126 - 126 - 126) nicolas cellier 05-05-06 22:10 edited on: 04-30-07 19:07	This bug is solved at 0003488

Issue History
Date Modified	Username	Field	Change
03-29-06 01:28	nicolas cellier	New Issue
05-05-06 22:10	nicolas cellier	Note Added: 0004884
04-30-07 19:07	nicolas cellier	Note Edited: 0004884
08-21-10 13:11	nicolas cellier	Relationship added	child of 0007002

Mantis 1.0.8[^]

40 total queries executed.
29 unique queries executed.