OASIS eXtensible Access Control Markup Language (XACML) TC

 View Only

Re: [xacml] RE: Your input needed on Comment#33. Forwarded messagefrom Daniel Engovatov.

  • 1.  Re: [xacml] RE: Your input needed on Comment#33. Forwarded messagefrom Daniel Engovatov.

    Posted 11-22-2002 13:10
     MHonArc v2.5.2 -->
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    

    xacml message

    [Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


    Subject: Re: [xacml] RE: Your input needed on Comment#33. Forwarded messagefrom Daniel Engovatov.


    
    The map function is polymorphic.
    
    The type of the map function is deduced by the function parameter and the
    other argument has to coincide.
    
    The type of map is  ( ( a -> b ) -> [a] -> [b] ).
    
    That is, map takes a function that transforms an element of one type to
    another, a bag of the first type, and returns a bag of the resultant type.
    
    What would "integer-map" mean? Would that mean a transformation of a bag
    of integers to integers? Somehow, that would defeat the purpose of the
    brevity of the specification, especially if you wanted to convert doubles
    to integers, or dates to strings, etc.
    
    If you want to define explicit map functions you would have to use the
    cross product of types, i.e. *-*-map. You define an awful lot of
    functions. Any good type checker can handle this. Also that would limit
    the use of the function to just the primitive types. If you invent another
    type, you have to define a new type-*-map and *-type-map functions, (at
    least for every *type you you have transformation functions defined. That
    would defeat the beauty of higher order functions, reducing the
    specification by extracting common functionality.
    
    As far as extension APIs. I whole heartily disagree with Daniel. I do care
    about them. I dont have any problem in defining new types and applying the
    higher order functions to any well defined function.  Maybe this is just
    an implemenation issue.
    
    -Polar
    
    
    On Fri, 22 Nov 2002, Anne Anderson wrote:
    
    > ------- start of forwarded message -------
    > From: Daniel Engovatov <dengovatov@crosslogix.com>
    > To: "'Anne.Anderson@Sun.com'" <Anne.Anderson@sun.com>,
    >    Daniel Engovatov
    > 	 <dengovatov@crosslogix.com>,
    >    Polar Humenn <polar@syr.edu>
    > Subject: RE: Your input needed on Comment#33
    > Date: Thu, 21 Nov 2002 14:59:55 -0800
    >
    > I would second this motion - I suggested it some time ago, but somehow it
    > got lost.
    >
    > It is indeed redundant to some extend, but for consistency it should be done
    > I believe.
    >
    > Now that we allow function to be a parameter, having a top level funciton of
    > undefined type really complicates proper extension API design.   But nobody
    > seems to care about extensions at this point.  Well, eventually you will,
    > but it would be too late to fix it.
    >
    >
    >
    >
    >