[MOBY-l] Boolean object

[Paul Gordon]

Wikipedia to the rescue! :-)

http://en.wikipedia.org/wiki/Canonical#Computer_science

In CS, "canonical" is more like "best-practice convention" than 
canonical in the religous or mathematical sense.  "true" and "false" are 
"best-practice" lexical representations for booleans.  Of course, since 
MOBY's lexical and semantic rules stand alone (though they could be 
*expressed* in XMLSchema, DTDs, Relax-NG, etc.), there is no requirement 
to allow 0 and 1 as valid values.  I realize that I'm being pedantic, 
but I think that giving people choices on how to express such a simple 
notion brings unneeded (albeit minor) complexity to the API.

My $0.02 :-)

P.S.  One advatange of 0/1 is that it is not specific to English.  But 
given that the entire MOBY API is in English anyway...

Gary Schiltz wrote:

> I think I'm in Paul's camp here, although I notice that the W3C 
> recommendation on XML Schema allows true, false, 1, and 0 as valid 
> lexical values for booleans (see www.w3.org/TR/xmlschema-2/#boolean). 
> But then, they say that the "canonical representation" for booleans is 
> true and false. I am still confused as to the difference between the 
> lexical and canonical representations.
>
> // Gary
>
> Paul Gordon wrote:
>
>> May I suggest that the boolean primitive have the one of the values 
>> "true" or "false", not "0" or "1".  This would alleve any confusion 
>> with integers, is more human readable, and is valid since "0" and "1" 
>> in Boolean logic are only arbitrary symbols in a lattice system (i.e. 
>> it is the union, intersection and negation functions that define the 
>> logic, not the symbols used).
>
>
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>