PR# 18503 Semantics of equality does notmatch the standard

Problem Report Summary

Submitter: alexk_es

Category: EiffelStudio

Priority: Medium

Date: 2013/02/04

Class: Bug

Severity: Serious

Number: 18503

Release: 7.2.9.587

Confidential: No

Status: Analyzed

Responsible:

Environment: Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Win64; x64; Trident/5.0)

Synopsis: Semantics of equality does notmatch the standard

Description

According to the standard, equality of expanded types means a result of is_equal. It does not say that the types need to be the same. As a result it's possible that (a=b)/=(b=a) if types of a and b are different and define is_equal in a different way.
This may need to be fixed in the standard, but so far this is a bug in the implementation.

To Reproduce

test#exec353

Problem Report Interactions

From:alexk_es Date:2013/02/08 Download

According to the standard

    The Boolean_expression e ~ f has value true if and only if
    the values of e and f are both attached and such that
    e.is_equal (f) holds.

 It does not say anything about type conformance. I guess the standard needs to
 be updated, but would like to keep this bug report until this is fixed.

From:manus_eiffel Date:2013/02/07 Status: Analyzed Download

I thought that = was ~ for expanded types, thus it cannot compare different types.