> The empty list is also a list of cheeses. Further down that road, I was a tiny bit surprised recently to discover that `andmap` returns true for any predicate when the input list is empty [1], but then I saw how it's a natural consequence of this principle. Empty is a member of every list, thus empty can never test false.
[1] http://docs.racket-lang.org/reference/pairs.html#%28def._%28%28lib._racket%2Fprivate%2Fmap..rkt%29._andmap%29%29 On Sat, Jan 3, 2015 at 11:34 AM, Jens Axel Søgaard <jensa...@soegaard.net> wrote: > 2015-01-02 23:58 GMT+01:00 Frank Weytjens <fweytjens.andr...@gmail.com>: > > Happy new year Racketeers, > > > An empty list is a list, but is it a lat ? > > > If you give an empty list as argument to the lat?-function, the answer is > > yes. But an atom must be a string of characters or numbers, or even one > > character, or a combination of special characters as long as it is not > an ( > > or an ). The empty list contains no characters at all, so it can not be a > > list of atoms. > > > Then an empty list can be anything. > > To some degree it is a matter of convention. This illustrates the > usual convention: > > Is this is a list of (person) names? > Alice, Bob, Charlie > Yes, all items on the list are names. > > Is this a list of names? > Alice, Bob, Cat > No, since Cat is not a name. > > Is this (the empty list) a list of names? > <nothing> > Yes, there are no non-names on the list. > In other words all items on the list are names, which makes it a list of > names. > > The empty list is also a list of cheeses. > > > -- > Jens Axel Søgaard > > ____________________ > Racket Users list: > http://lists.racket-lang.org/users >
____________________ Racket Users list: http://lists.racket-lang.org/users
