On Thu, 18 Sep 2025 04:44:28 GMT, Chen Liang <[email protected]> wrote:
>> Add a total that the total order used by {Double, Float}.compareTo is
>> different than the total order defined by IEEE 754, starting the 2008
>> version of that standard.
>
> src/java.base/share/classes/java/lang/Double.java line 1443:
>
>> 1441: * that is NaN representations whose sign bit is set, to be less
>> 1442: * than any finite or infinite value, including negative
>> 1443: * infinity. Also in the IEEE 754 ordering, "positive" NaN values
>
> This last sentence seems redundant because the behavior for positive NaN
> values are the same for both total orders.
Hmm. Yeah, I wanted to avoid giving too much details on what IEEE 754 does
here; they also ascribe ordering between different NaN bit patterns while
Double.compareTo puts all NaNs into the same equivalence class. I'll take
another run at the text; thanks.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/27356#discussion_r2360263378
