Re: bignum branch

[Andy Moreton]

On Wed 18 Jul 2018, Clément Pit-Claudel wrote:

> On 2018-07-18 09:18, Stefan Monnier wrote:
>>>> For example, (eql 0.0e+NaN -0.0e+NaN) now returns t
>>>> because the two NaNs have different signs; formerly it returned nil.
>>> Should this be the other way around ?
>> You mean it's (eql -0.0e+NaN 0.0e+NaN) which should return t?
>> Shouldn't `eql` be commutative?
>
> No, I mean that Paul wrote 't' where he meant 'nil' and 'nil' where he meant 
> 't'.

I agree:

    ELISP> emacs-repository-version
    "c70d22f70b77b053d01c7380122d166ecb728610"
    ELISP> (eql 0.0e+NaN 0.0e+NaN)
    t
    ELISP> (eql -0.0e+NaN -0.0e+NaN)
    t
    ELISP> (eql 0.0e+NaN -0.0e+NaN)
    nil
    ELISP> (eql -0.0e+NaN 0.0e+NaN)
    nil