On 2012-01-12, at 17:25, John Cowan wrote:

> Vincent Manis scripsit:
>
>> How about `Rational operations such as + should always produce exact
>> results where possible when given exact arguments'?
>
> On reflection, I think the definition of SHOULD does cover this.
> "There may exist valid reasons in particular circumstances [for an
> implementation] to ignore a particular item", namely that it does not
> allow arbitrarily large exact numbers.

It doesn't feel like it covers it to me; that sentence and the next seem
contradictory (it's an apparent, not a real, contradiction). But the
animal cruelty people have been on at me about flogging dead horses :)

-- vincent

_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports