> > They should both [pointers and allocatables] remain, I was thinking
> > about obsoleting certain USES of them (but thinking about it, don't
> > know if that is practical or can even be done).
> >
> > GOTOs might still have their uses, but emulating IF blocks and DO loops
> > is not one of them.
>
> Ok. Agree in a way. But I don't think you are talking about
> something in the standard. I can't picture any way for the standard
> to say anything like that. The legality of a construct can't very
> well depend on a user's intent.
Right. I was thinking that there might be cases where it is obvious
that a pointer is being used because (currently) an allocatable is not
allowed, but that might not be the case.
> Sounds to me more like you are making suggestions for something like a
> style guide. As such, I think its a fine idea. That is as long as it
> did not use terminology that could be so easily confused with a
> statement about the standard.
Right. The term "obsolete" is, since F90, now standardese, I suppose.
> I don't like it when readers of a book
> come away thinking that style choices suggested by the book are a
> mandate of the standard. I don't like this kind of confusion even
> when I agree with the style choices in question.
Couldn't agree with you more here. I remember your flames of a
particular book on this count. :-)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|