Re: why doesn't this compile ?

In article <-KudnfV6vpCniOzfRVn-jw@xxxxxxxxxxx>,
"James Van Buskirk" <not_valid@xxxxxxxxxxx> wrote:

> "Richard E Maine" <nospam@xxxxxxxxxxxxx> wrote in message
> news:nospam-BF37BA.08405228042005@xxxxxxxxxxxxxxxxxxxxx
>
> > (and in that case, it is
> > restricted to being a module procedure just because it is strange to
> > have a "module procedure" statement refer to something that isn't).
>
> Why didn't you just standardize existing practice and permit
> any procedure whose interface is explicit to be named as a
> module procedure?

I'm not sure whether I understand your question.

First, I wasn't aware that it was existing practice to allow procedures
other than module procedures to be specified in a module procedure
statement. The standard's prohibition against this is a constraint, so
existing compilers are required to be able to at least diagnose it. If
they can't that is a compiler bug. I don't know how many current
compilers might allow this (nag doesn't, g95 does) and how many of those
that do allow it might soon get fixed to at least be able to diagnose
it. I've never happened to see this one before (which means I had not
seen it at all when this f2003 feature was being debated), so I'm
dubious about labeling it as "existing practice". Even if it was
"existing practice", I for one, was not aware of this at the time (and
I'm not yet convinced), which is probably at least a partial answer to
the question. Maybe not an answer you'd like, but it would still be one.

Second, I thought I had pretty explicitly answered the question with my
parenthetical remark quoted above. Whether one agrees with that as a
good reason is a different matter from whether it is a reason. It was,
in fact, my reason for agreeing with the decision at the time. This
leaves me thinking that either

1. I misunderstand exactly what you are asking.

2. You overlooked or didn't understand my remark above.

3. Your "question" was actually a statement of disagreement phrased as a
question, in which case I'll just accept the disagreement without
further comment.

--
Richard Maine | Good judgment comes from experience;
email: my first.last at org.domain | experience comes from bad judgment.
org: nasa, domain: gov | -- Mark Twain
.

Relevant Pages

  • Re: generic module procedures with function arguments
    ... even when using strict standard adherance compiler switches. ... > module procedure GQC ... is there any way to correct to be standard conforming ... I tried with 3 compilers: ...
    (comp.lang.fortran)
  • Re: public/private and overloaded operator
    ... module procedure is of a type that has PRIVATE accessibility and the ... some sense, both compilers are right. ... particular compiler options for controlling the version of the standard ...
    (comp.lang.fortran)