(j3.2006) GB44 - delete A.1.3

[Malcolm Cohen]

Bill Long wrote:
>Comment GB44 wants to delete subclause  A.1.3.  This has been proposed in the 
>past and been rejected.

Somewhat beside the point.  Country comments have somewhat greater significance 
than those of J. Random Commitee-Member.

>  The Comment seems to miss the main point of the example - it is the only 
> example given that illustrates the use of TYPE(*) in an actual Fortran 
> subprogram.

Except that it does not.  Not even a little bit.

>  Something I would have thought obvious with the leading line starting 
> "Example of TYPE(*) ...", but I guess not.

Heading does not an example make.

>   The point of the example is not to illustrate the use of the generic 
> interface.

But it is the only thing it does.

>Instead of deleting the only illustration of this new functionality, I would 
>propose fixing the one problem that (assuming I can guess what "64-bit-safe" 
>means) is raised.  Perhaps adding "For implementations for which C_int and 
>selected_int_kind(17) have different values...".  Which is just about every 
>system I know of that will have a Fortran 2008 compiler.

Replacing it with an *example of its use in a Fortran program* would be a better 
step in the right direction.  Giving a C prototype and a matching Fortran 
interface is completely trivially derived from the syntax and a waste of ink.

The generic interface is likewise a waste of ink and a netful of red herrings to 
boot.

>This example was included because it solves a nagging problem caused by the 
>vile -i8 compiler switch, and I would expect this construct to be frequently 
>used.

This has ABSOLUTELY nothing to do with TYPE(*) or anything else in the TR 
whatsoever.  Those of us writing C interfaces have been doing this for ages, and 
even more since F2003 came out.

I am not personally responsible for this UK comment but I certainly agree with 
it.  Delete A.1.3, or throw it away entirely and produce a new example that 
actually shows TYPE(*) usage without clever discourses into showing other tricks 
that are not new but distractions from whatever point A.1.3 should have been 
making.

Cheers,
-- 
................................Malcolm Cohen, Nihon NAG, Tokyo.