RH = Roger Hui, DB = me.

RH>  What can you do in that time?  

Again, you're using your definition of "have", not mine.  So our discussion is 
moot.

I said:

DB>  I define "have" to mean "the bit pattern for an IEEE floating point
DB>  (FP) negative zero exists in an array of FP in J".  

To satisfy this definition, the negative zero need only exist in J.  The 
exploits I laid out meet that criterion.  

RH>  I can also use part of my "one hour" to disable all 
RH>  15!:x foreigns.

Sure, or you could spend 15 seconds adding the line "A J array cannot have a 
negative zero (if the bit pattern of a IEEE floating point negative zero is 
introduced into a J FP array, it is no longer a J array and the interpreter's 
behavior from then is no longer covered by this Dictionary)" and spend the 
balance of the hour having a nice lunch.

But that invalidates the premises of the discussion.  The meaning of the term 
"J" would have changed, and so my comments may no longer apply.

You cannot win.  So long as it runs on a machine implementing the now-current 
IEEE floating point standard, J will have negative zeros.

-Dan

PS:  I really need to stop playing on the Forums so much. I nearly missed a 
meeting because of this discussion!  And I was hosting the meeting!
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to