(j3.2006) derived types with type parameters are different

[Robert Corbett]

Jim Xia wrote:
> 
> j3-bounces at j3-fortran.org wrote on 07/25/2008 12:44:06 PM:
> 
>  >
>  >
>  > Robert Corbett wrote:
>  > >
>  > > Are there any compilers that implement parameterized derived types
>  > > yet?
>  >
>  > Yes, and I would expect that some that don't have at least gotten past
>  > the design stage.  The feature is a pain to implement, so resistance to
>  > doing it over a different way will be very high.
> 
> We're still implementing it.  As a general comment on public comment 
> J32004 "F2008 too early": there is a call to review F2003 and address 
> the reason for lack of implementations.  Aside from the size of F2003, 
> there are also features not well designed.  Parameterized derived type 
> is one of those falls into this category.
> 
> 
>  > > If so, how do they handle type equivalence?
> 
> At the moment our compiler handles N vs. N+0 case, and we reject N+N vs. 
> N*N case.

What about the 2*N vs. N+N case?

Bob Corbett