Re: boolean datatype ... wtf?
From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Wed, 29 Sep 2010 15:42:52 -0300
Message-ID: <4ca388a8$0$14803$9a566e8b_at_news.aliant.net>
> ...
>
>
> Yes, I was taught that dividing the integer 1 by the integer 0 is
> undefined. Maybe I should have suggested such an op be called
> 'relational integer division'. Surely the set of integers for which 1/2
> results in an integer is empty?
Date: Wed, 29 Sep 2010 15:42:52 -0300
Message-ID: <4ca388a8$0$14803$9a566e8b_at_news.aliant.net>
paul c wrote:
> On 29/09/2010 10:04 AM, Bob Badour wrote:
>
>> paul c wrote: >> >>> On 29/09/2010 8:45 AM, Bob Badour wrote: >>> >>>> paul c wrote: >>>>
> ...
>
>>> Regarding division by zero, as impractical as it might be, I also >>> might have a mental block about that. If there were an op called >>> 'integer division', is there a theoretical reason why its operands >>> couldn't be relations and the result an empty relation? >> >> No. The result of 1/0 is undefined not an empty set of values, and the >> result of 0/0 is indeterminant not the set of all integers. Dividing by >> zero is not the only case where division is not closed on integers: 1/2 >> is not an integer.
>
> Yes, I was taught that dividing the integer 1 by the integer 0 is
> undefined. Maybe I should have suggested such an op be called
> 'relational integer division'. Surely the set of integers for which 1/2
> results in an integer is empty?
But the result of division is not empty. Received on Wed Sep 29 2010 - 20:42:52 CEST