Opened 6 years ago
Closed 5 years ago
#314 closed defect (duplicate)
feature review
Reported by: | alexshinn | Owned by: | alexshinn |
---|---|---|---|
Priority: | major | Milestone: | |
Component: | WG1 - Core | Keywords: | |
Cc: |
Description
The list of standard features needs more careful review.
In particular, exact-complex seems odd - it's more likely
that we want to test for any complex, and if we really want
to distinguish exactness why not inexact-complex?
Change History (3)
comment:1 Changed 6 years ago by cowan
comment:2 Changed 6 years ago by cowan
Okay, back home. Support for all types of complex numbers: MIT, Gambit, Mosh, STklos. (Also CLISP.)
Support for exact/exact and inexact/inexact complex numbers only (no mixed exactness): Racket, Chicken (with the numbers egg), Scheme48/scsh, Kawa, Chez, Vicare, Larceny, Ypsilon, IronScheme. (Also Armed Bear CL, CMU CL, Clozure CL, ECL, GNU CL, Steel Bank CL.)
Support for inexact/inexact complex numbers only: Gauche, Guile, SISC, Chibi, SCM, KSi, Scheme 7, UMB.
No support for complex numbers at all: Bigloo, Ikarus, SigScheme, RScheme, Elk, VX, Oaklisp, Owl Lisp.
comment:3 Changed 5 years ago by cowan
- Resolution set to duplicate
- Status changed from new to closed
Equivalent to #320.
I don't know of any Schemes that support exact complex numbers but not inexact. The normal cases are all types of complex numbers, or inexact-inexact complex numbers only (like SCM and some others -- I don't have access to my suite of Schemes today). So the exact-complex test is for the ability to do exact arithmetic on complex numbers. To test for the presence of complex numbers of any kind, test for (module (scheme complex)).