list algebra
From: mAsterdam <mAsterdam_at_vrijdag.org>
Date: Tue, 29 Jun 2004 22:03:32 +0200
Message-ID: <40e1cb0b$0$36861$e4fe514c_at_news.xs4all.nl>
>
> Manipulate lists. One would have a set of functions that take
> lists as arguments and return lists as results. The goal in my mind
> is to have something that would be reminiscent of the completeness
> criteria that one can use to evaluate set algebras.
>
>
> String is a subtype of list, where the element type is character.
> List is a subtype of tree, where each node in the tree has only one child.
>
>
> The lambda calculus deals with functions. Operations include
> abstraction and application. A list algebra would deal with
> lists. Operations would include things like add and delete.
> There would be no application operation. Also, the lambda
> calculus is a calculus, whereas my list algebra would be an algebra.
>
>
> Any list.
Date: Tue, 29 Jun 2004 22:03:32 +0200
Message-ID: <40e1cb0b$0$36861$e4fe514c_at_news.xs4all.nl>
Marshall Spight wrote:
> x
>>What do you plan to do with this algebra of lists ?
>
> Manipulate lists. One would have a set of functions that take
> lists as arguments and return lists as results. The goal in my mind
> is to have something that would be reminiscent of the completeness
> criteria that one can use to evaluate set algebras.
>
>>How your list is different from a string or a tree ?
>
> String is a subtype of list, where the element type is character.
> List is a subtype of tree, where each node in the tree has only one child.
>
>>How your algebra would be different from lambda calculus ?
>
> The lambda calculus deals with functions. Operations include
> abstraction and application. A list algebra would deal with
> lists. Operations would include things like add and delete.
> There would be no application operation. Also, the lambda
> calculus is a calculus, whereas my list algebra would be an algebra.
>
>>You plan to use any lists or some particular kind of lists ?
>
> Any list.
Heterogeneous lists look like they can create a lot of (for the order problem unnecessary) complications when looking at them in a relational framework. Homogeneous lists (all elements of one type only) would suffice for the 'order' problem sec - or am I just seeing ghosts? Received on Tue Jun 29 2004 - 22:03:32 CEST