(j3.2006) a question on same_type_as and type parameters

[Aleksandar Donev]

On Thursday 28 August 2008 07:55, Bill Long wrote:
> I don't think this reference is legal.
OK, that covers the intrinsic and SEQUENCE/BIND(C) case.

> The "same type" wording relies on the description of what is
> means for two derived types to be the same.
Can you point me to those words. Also, answer Jim's second question: Are 
type(t(5)) and type(t(10)) the same type? My reading seems to indicate that 
yes, they are the same type.

Perhaps it would have been better to say "same type and kind type parameters", 
but maybe there is a reason we did not say that?

My firm belief is that this whole business about type parameters was mostly a 
mistake. Parameterized types/codes should be done using something like 
templates/macros, i.e., in a pre-processing step, rather than built into the 
language like this. It causes big implementation problems, lots of tricky 
cases, and most of all, the main issue, that of writing routines to go along 
with the types, is unresolved.

Best,
Aleks