Version française
Home     About     Download     Resources     Contact us    

This site is updated infrequently. For up-to-date information, please visit the new OCaml website at

Browse thread
[Caml-list] Is a Cow an Animal?
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: 2002-04-19 (00:51)
From: james woodyatt <jhw@w...>
Subject: Re: [Caml-list] Is a Cow an Animal?
On Thursday, April 18, 2002, at 05:21 PM, Remi VANICAT wrote:
> james woodyatt <> writes:
>> I borrowed a technique I learned here from Brian Rogoff that uses an
>> abstract type with a contravariant type parameter for passing the type
>> of energy a food object contains to an animal object that can eat it.
>> (I wish I were smart enough to know the name for this technique.)
> the use of an abstract type with a covariant or contravariant type (or
> even not variant) parameter, and where the actual implementation of
> the type doesn't use this parameter (as in type 'a eater = int) is
> often call phantom type.
> one can read the very interesting mail about this in the archive of
> this mailing list

That's precisely where I learned the technique.  It's true I forgot the 
"phantom type" terminology, but the contribution I was trying to make to 
the list is that such types are frequently useful in representing 
complicated associations between related class hierarchies.

I don't think Pixel's exercise is really covered well by the 
"subject-observer" pattern, and I wonder if the pattern I used in my 
solution is one that is already identified by a better name than the 
"association-by-phantom-type" pattern.

Surely somebody has already published a paper on this by now, right?

j h woodyatt <>

To unsubscribe, mail Archives:
Bug reports: FAQ:
Beginner's list: