Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: Division in Draft Standard Text, V02.1



"Prof.Dr. Jürgen Wolff von Gudenberg" wrote:
John, P1788

We have some suggestions to sharpen the definition of division by
zero. These may be considered in the light of defining a simple
hardware standard, and as a reaction to the comments on motion 11.

the current wording of the (draft) standard contains several
suggestions for division by B with b1<0<b2
the result of A/B is defined as [-oo,+oo] and sets trits: defined =-1,
continuous=-1

the user who wants to receive the 2 parts of the set result has to
call A/[b1,0]  and A/[0,b2] separately.
A language implementation could easily define a function doing this.
(see C++ draft for interval class)

discussion open

Juergen and Marco


Juergen,

What you explain above is my preference and recommendation, as well.

Authors of Computer Algebra Systems or sophisticated class libraries can provide support for multi-intervals, but I'm not sure P1788 should specifically incorporate this concept. It will be very difficult for hardware.

Nate