Re: [ublas] Generic traits for scalars

From: Dima Sorkin (dsorkin_at_[hidden])
Date: 2005-07-11 02:30:01

Just forgot to add:
Quoting Andrew Rieck <arieck_at_[hidden]>: 
> Your model of scaler_traits is really a fat interface for a complex type, 
> not for a generic scaler_trait. 
> There is no need to include member functions like abs, etc, in a traits 
> interface. 
 I had problems in past with definitions of "abs" etc, so I don't think
that the interface is too fat. If you have some "scalar" class where
"abs" is called "norm" , you may prefer to be inside the frame of a standard
traits class, then to write your own adaptor.
 The clear disadvantage is that you will have to declare things twice,
if you already have "abs"-semantics function with the right name.
  It is only the question of making a decision of how the interface will
look like.

Regards,
 Dima.