[bitc-dev] BitC Overloading
Jørgen Hermanrud Fjeld
jhf at hex.no
Fri Jan 28 12:24:41 EST 2005
On Fri, Jan 28, 2005 at 07:43:27AM -0800, Shawn Garbett wrote:
> Does avoiding polymorphism and static type checking
> exclude things like type classing in Haskell?
As far as I know you can type check type classes statically, but there
is run time overhead because of type information. The work on
extensional polymorphism offers another solution instead of type
classes, and with better contol over the overloading.
Extensional polymorphism also has a run-time representation of
types, and use flow-analysis to optimise this representation.
If you have static type checking, you can add polymorphism and type
classes or extensional polymorphism at a later time. Extensional
polymorphism was developed as a constraint layer on top of the O'Caml type
> Type classing in Haskell is a very elegant solution to
> developing generic algorithms and doesn't bring with
> it all the problems that OO polymorphism has. However,
> I can understand why this would cause problems with
> theorem provers and other kernel runtime issues.
Some of the samples with extensional polymorphism can type more than
possible with type classes.
Sincerely | Homepage:
Jørgen | http://www.hex.no/jhf
| Public GPG key:
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: Digital signature
Url : http://www.coyotos.org/pipermail/bitc-dev/attachments/20050128/2f0b2569/attachment.bin
More information about the bitc-dev