(j3.2006) (SC22WG5.5161) [ukfortran] [ Draft corrigendum 3]

[Malcolm Cohen]

Bill Long writes:
>Reason for No vote:
>---------------------------
>
>I agree with Malcolm's original comment that Interp f08/0086 should be 
>failed and restarted.

Right, originally I did not think I would have the time to work on it.  I 
made the time available.

> I'm not happy with the proposed last-minute text file of changes for these 
> reasons:
>
>1) A main reason for having a vote like this is to agree (or not) on the 
>exact form of the PDF file form of the Corrigendum - this is basically an 
>"FCD" vote.  That purpose is subverted here because the proposers of this 
>path have not supplied the PDF form of their replacements.

Neither have you supplied PDF for your proposed changes.  It might be 
considered that any change to the PDF beyond the utterly trivial should 
incur a re-vote, but that does not seem to be what you are arguing.

We have corrected mistakes in the edits (though more trivial than this one) 
in previous corrigenda votes, so I don't think this is a time for voting NO 
on principle.  (The time for discussion of the principle was, in my opinion, 
when I first suggested that we actually fix it, rather than now, after we've 
done the work...)

>2) Another goal of a final vote like this is to reflect on whether we have 
>made a technical mistake.  The complexity of the changes proposed for 
>f08/0086, and the problems getting them "right" convince me that we chose 
>the wrong answer for the interp.

You are correct in that no change to the answer is being proposed.

>  The alternate answer would result in a trivial, single edit in the 
> standard, and  would prohibit only code like that in the interp example 
> that is confusing at best and would never be allowed by any reasonable 
> coding style guide.   The place to have this discussion is in the process 
> of restarting the interp.

We already had precisely this discussion, and it failed the J3 letter 
ballot, that is how we got here.  The exact proposal you described was voted 
NO by 5 people (a majority) on the J3 ballot.

I (and others) disagree with your assertion that
   REAL x(*)
   PARAMETER(x = [ 1,2,3,4,5 ])
is unreasonable and should never be allowed.  Disallowing this would break 
one of our basic design rules for the BNF of declarations, which is that one 
can specify attributes independently.  Yes it is a smaller edit to the 
standard, and would not remove significant functionality ... but then 
deleting the PARAMETER statement entirely would not remove significant 
functionality!

If we did the small edit to the standard you suggest, one might conjecture 
that one or more of the previous 5 NO-voters would immediately raise this 
violation of our basic design rules as a wart, and get it fixed in F2015! 
Rendering the whole "trivial single edit" idea a waste of time.

Cheers,
-- 
...............................Malcolm.