Promotion inconsistency in min/max pair float

[hiiamboris]

Describe the bug

>> min 1x1 (1.#inf,1.#inf) 
== (1, 1) 
>> min 1x1 1.#inf 
== *** Script Error: integer! overflow/underflow
*** Where: min
*** Near : min 1x1 1.#INF

These operations are reasonable to expect to be equivalent, but they aren't. Also:

>> max 1x1 (1.#nan,1.#nan) 
== (1.#NaN, 1.#NaN) 
>> max 1x1 1.#nan 
== *** Script Error: integer! overflow/underflow
*** Where: max
*** Near : max 1x1 1.#NaN
*** Stack: startup process-into process-into thrown call-cli-func do-events trap with-thrown process-any-event process-next-event dispatch with-stop process-event do-handlers do-handler trap with-thrown evaluate-since evaluate trap with-thrown capture-output fcatch with-thrown 

Expected behavior

Promote scalar into relevant 2D type then compute min/max.

Platform version
Red 0.6.5 for Windows built 22-Feb-2024/16:47:26+03:00 commit #5521b51