[r6rs-discuss] [Formal] ->exact and ->inexact should be renamed

[R. Kent Dybvig]

---
This message is a formal comment which was submitted to formal-comment at r6rs.org, following the requirements described at: http://www.r6rs.org/process.html
---
name:      Kent Dybvig
email:     dyb at cs.indiana.edu
type:      defect
priority:  minor
component: arithmetic
version:   5.92

Summary: ->exact and ->inexact should be renamed

Description:

These names ->exact and ->inexact are ugly.  They also vaguely imply a
conversion is necessarily occuring and, therefore, that ->whatever should
not be applied to something that is already whatever.  (They are, however,
better in this regard than the original names inexact->exact and
exact->inexact.)

Proposal:

Change the name ->exact to exact and change the name ->inexact to inexact,
so that, e.g., (exact 3.0) and (exact 3) are both 3 and (inexact 3.0) and
(inexact 3) are both 3.0.

These names appear more or less like casts and do not imply that a
conversion is necessarily taking place.  They are also consistent with the
names char-upcase, string-downcase, etc., which contain no -> and also
name procedures that operate on values that may need no conversion.  To my
mind, they also seem more consistent with the #e and #i exactness
prefixes.