Ken Dickey wrote:
> On Tuesday 21 October 2008 14:57:19 you wrote:
>
>>> should read:
>>> Ken's definition => "a singleton or empty list is _not_ ordered".
>>>
>>>
>> And you should be saying "monotonic" or "sorted" as well.
>>
>
> Precisely. Something that cannot be compared with cannot be ordered or
> sorted
> so cannot be monotonic [IMHO].
>
> Now for (N < 2) you could return a (prepositional) function which would take
> further numbers and become a preposition which would return a boolean result
> (or raise an exception or halt the machine or ...). You might think of a
> binary relation which starts out with a negative number of arguments and
> returns (values <OK-so-far-OR-more-needed?> <next>), where you could call
> <next> with further arguments and would get another value pair. If there
> were not enough arguments (e.g. one argument to a binary predicate)
> then 'more-needed (or some such marker) would be the first result. If two or
> more arguments, the first result would be a boolean indicating if the result
> so far was monotonic/sorted -- so far.
>
> That would make more sense to me than returning a boolean result to an
> incomplete question.
>
> In a conditional test, multiple values could be accepted and only the first
> one used.
Right.
> Somehow I suspect this is a more interesting/complex
> computational model than most Scheme implementers or users are interested in.
>
How can we change that? I wish I could go back in time and
pose that question at the start of this discussion.
But, back on earth:
The clear implication of all of the above is that we would grow
as a community of knowledge better if we could agree
1) < > = etc. should be strictly arity-2 procedures in the core
2) let's see what libraries people write, given that, now that we
have modules
-t
> Cheers,
> -KenD
>
>
>
_______________________________________________
r6rs-discuss mailing list
[email protected]
http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss
