LCAS comments, and X3T2 proposal for changes to LCAS <9201251510.AA22235avalidgh.com>

[queueing-rmail]

David G. Hough writes:
  
  A much more serious problem for mathematical software portability is 
  implementations that claim IEEE 754 compatibility because they use the
  storage formats, the least important part, and ignore or implement poorly
  the more important parts of the standard requiring correctly-rounded
  multiplication, division, square root, easy base conversion cases,
  and correct handling of subnormal operands and results.

Just out of curiosity, does anyone know if there exists anywhere any
sort of a test suite for IEEE 754 conformance?  Having one written
in C would be nice, but I'd like to know if there is any such a thing
written in any language.


// Ron ("Loose Cannon") Guilmette
// Internet: rfgancd.com      uucp: ...uunet!lupine!segfault!rfg
// New motto:  If at first your don't succeed, forget it.