13:34:25scymtymok, that's kind of a big deal for correctness as well as performance
13:34:26jackdanielwhich does "the right thing" (i.e in case of real it just returns n and is declared inline)
13:34:46jackdanieland in case of the previous representations it does indeed coerce
13:35:03scymtymso what is "the right thing" in terms of the spec? REAL?
13:35:32jackdanielin terms of spec there is no the right thing, implementations are allowed to approximate with floats but they doesn't have to do that
13:35:53jackdanielas I said, I don't remember which chapter it's stated in, just remember the statement itself
13:36:35scymtymCOORDINATE is currently the identity, if i'm not looking at the wrong version
13:46:50scymtymit does :x1 (min x1 x2) :x2 (max x1 x2)
13:47:13scymtymwhat is the keystroke specifier for the f3 key?
13:47:21jackdanielyes, in the instance, that's why they are called. I'm saying that make-bounding-rectangle arguments does not have this property guaratneed
13:48:12scymtymMIN and MAX are applied to the same arguments (for each dimension)
13:48:26scymtymso two comparisons (for each dimension)
13:48:33scymtymbut only one is needed (for each dimension)
13:49:18jackdanielso you want to put there two <= (one for each dimension)?
13:49:48jackdanielif so, then I misunderstood what you said