(j3.2006) (SC22WG5.5126) [ukfortran] [ Draft corrigendum 3]

[Malcolm Cohen]

J3 people, please remember this discussion is on the WG5 list and change the 
address when you "reply".

Thanks to Van for spotting the typo.  Yes, "<implied-assumed-spec>" (in the edit 
for [95:33-]) should have been "<assumed-implied-spec>" (as you might have 
guessed, the name of the new syntax term went through a couple of revisions, and 
I missing changing one of the earlier versions).

Van writes:
>it seems
>that the revised R522 at [96:26] results in not being able to declare
>the bounds for a rank-one array."

See the immediately preceding edit, already in the previous version, which says
[96:24-25] 5.3.8.6p1
  Replace sentence
    "An implied-shape array ... <implied-shape-spec-list>."
  with
    "A named constant is declared to be an implied-shape array by an
     <implied-shape-or-assumed-size-spec> or an <implied-shape-spec>."
{Now two ways of declaring implied shape.}

<implied-shape-or-assumed-size-spec> is the ambiguous "might be assumed-size" 
rank-one case.  <implied-shape-spec> is the unambiguous rank>1 case.  So there 
should be no problem.

(BTW the edit for [96:26] in the interp was wrong, and the draft corrigendum 
attempted an incorrect fix, but I *think* the version in mjc002.txt works 
properly.)

Cheers,

-----Original Message----- 
From: Van Snyder
Date: ?? 25?12?5? 5:46
To: fortran standards email list for J3
Subject: Re: (j3.2006) (SC22WG5.5125) [ukfortran] [ Draft corrigendum 3]

In the attachment mjc002.txt to the original message, the edit for
[95:33-] defines a syntax term <implied-assumed-spec> but the text and
later syntax terms refer to <assumed-implied-spec>.

Maybe I haven't worked through all the details correctly, but it seems
that the revised R522 at [96:26] results in not being able to declare
the bounds for a rank-one array.

Since this is the last F2008 corrigendum, I don't object to correcting
the problems using ballot comments.

On Wed, 2013-12-04 at 15:27 +0900, Malcolm Cohen wrote:
> I wrote:
> > Please answer the following question "Is N1995, with the references and
> > notes removed, acceptable for submission to SC22 for publication as
> > Corrigendum 3 for Fortran 2008?" in one of these ways.
> >
> > 3) No, for the following reasons.
> >
> > The edits for F08/0086 do not work.  This interp must be failed and 
> > restarted,
> > and the corrigendum must be altered to remove it.
>
> Bill Long replied:
> >I'm happy with this comment.  But what is the larger plan?
> >
> >1) The rest of the Corrigendum is OK and your vote changes to YES if the
> >F08/0086 edits are removed and we move forward with what remains, or
>
> I have not checked the rest of the corrigendum yet...
>
> >2) Fixing the problem raised by F08/0086 is sufficiently important to
> >put the Corrigendum on hold until it is repaired and put up for another
> >vote, considering that this is the final Corrigendum for F2008.
>
> If we follow the normal "back to square one" procedure, that is a delay until
> April at the most wildly optimistic (February meeting + J3 letter ballot + WG5
> letter ballot).
>
> Here is a third alternative:
> 3) New corrected edits are unanimously agreed by WG5 and J3
>    during the Corrigendum ballot.  A single "No" vote should definitely
>    send this interp back to stage one (but not the rest of the corrigendum).
>
> I attach a corrected interp, the differences are:
>
> (i) New syntax term "<assumed-implied-spec>" (I did not like my own suggestion
> of "<star-bounds>").  "assumed-implied" rather than "implied-assumed" because
> assumed-size arrays are described first.
>
> (ii) I "improved" the edit for [95:33] cosmetically, as well as using the new
> syntax term.
>
> (iii) Used the new syntax term in R522, twice.
>
> (iv) New edit for [96:28] to use the new syntax term in the rank 
> determination.
>
> (v) Removed NOTE (1) and reformatted the remainder.
>
> I think this is right this time, but who knows?
>
> If anyone is unconvinced that this is right, or is uncomfortable with the idea
> of such a radical last-minute fix, we should send this back though the system 
> to
> get a more leisurely review.
>
> Cheers,
> _______________________________________________
> J3 mailing list
> J3 at mailman.j3-fortran.org
> http://mailman.j3-fortran.org/mailman/listinfo/j3


_______________________________________________
J3 mailing list
J3 at mailman.j3-fortran.org
http://mailman.j3-fortran.org/mailman/listinfo/j3

________________________________________________________________________
This e-mail has been scanned for all viruses by Star.
________________________________________________________________________

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