(j3.2006) (SC22WG5.4529) WG5 ballot on interpretations - N1876

[Bill Long]

Following are my (Bill Long, Cray Inc.) votes  for the interp ballot in 
N1876:

The following Fortran 2003 interpretations are being balloted:

Yes  No Number     Title
-Y- --- F03/0030   IEEE divide by zero
-Y- --- F03/0048   Control edit descriptors in UDDTIO
-Y- --- F03/0085   Finalizing targets of pointer or allocatable
-Y- --- F03/0091   Array components cannot depend on length type parameters
-C- --- F03/0096   Can a read statement change the unit value?
-Y- --- F03/0105   SIZE= specifier and UDDTIO
-Y- --- F03/0110   Restoring dropped restriction on ENTRY
-C- --- F03/0121   Precise FP semantics of the REAL intrinsic
-Y- --- F03/0123   Implicit typing in derived types
-Y- --- F03/0124   definition is poorly defined
-Y- --- F03/0128   Subobjects in namelist output
-Y- --- F08/0001   Generic resolution with pointer dummy arguments
-Y- --- F08/0002   Are assumed- or deferred-shape objects allowed in 
namelist?
-Y- --- F08/0003   Is a disassociated pointer allowed as an actual DIM
                    argument?
-Y- --- F08/0004   Is TARGET argument of ASSOCIATED a pointer or nonpointer
                    dummy?
         F08/0005*  optional arguments and ASSOCIATED - subsumed by F08/0004
-Y- --- F08/0006   generic resolution with banned argument combinations
-Y- --- F08/0007   Can zero have more than one bit sequence representation?
-Y- --- F08/0008   IEEE exceptions for intrinsic functions
-Y- --- F08/0009   Is ABS ever required to be the optional IEC 60559 abs?
-Y- --- F08/0010   deallocating objects that are associated with other 
objects
-Y- --- F08/0011   How many times are constructed values finalized?
         F08/0012*  Are constants finalized? - subsumed by F08/0011
-Y- --- F08/0013   How does finalization interact with allocatable 
assignment?
--- -N- F08/0014   Finalizing assignment to vector-subscripted object
-Y- --- F08/0015   IMPLICIT
-C- --- F08/0016   Can a vector-subscripted argument become undefined?
-C- --- F08/0017   Elemental subroutine restrictions
-Y- --- F08/0018   Impure elemental restrictions
-Y- --- F08/0019   Transformational Bessel functions
-Y- --- F08/0020   FINDLOC and logical arguments

Comments for C and N votes:

F03/0096:

The wording of the second edit suggests that the SIZE= value returned 
cannot depend on the input list, which is not what is intended.  I think 
this can be corrected  by adding " the values of"  after "shall not 
depend on" in the edit.

F03/0121:

My reading of the original question included a a desire to say that the 
result value for REAL(X, KIND=wp), for a particular value of X, is the 
SAME result value independent of the context in which the function 
reference appears.  After all, it would not really be a "function" in 
the normal sense if that were not the case.  The subsequent discussion 
in the ANSWER section could cast doubt on whether this basic requirement 
is actually true. It would be better to explicitly state that REAL for a 
particular set of arguments always returns the same result value 
independent of the context.

F08/0014:

The proposed edit is to add this text:  "A vector-subscripted array 
section shall not be finalized by a nonelemental final subroutine". 
This is stated as a requirement on the user (program).  It is unclear to 
me what this really means. Here are some options that come to mind:

1) For a type that specifies final subroutines, if a vector-subscripted 
array section of that type appears in a context where it would be 
finalized, the type shall specify an elemental final subroutine.

2) For a type that specifies final subroutines, if a vector-subscripted 
array section of that type appears in a context where it would otherwise 
be finalized, the section is not finalized if there is no elemental 
final subroutine.

3) For a type that specifies final subroutines, a vector-subscripted 
array section of that type shall not appear in a context where it would 
be finalized unless the type specifies an elemental final subroutine.

F08/0016:

In the third line of the ANSWER, there is a typo:  "copvers" -> ""covers".

F08/0017:

The discussion section paragraph:

"The subroutine test_add does not do anything useful other than to set 
the IEEE_OVERFLOW flag when x+y would overflow."

This is a bit of a shock, since almost all compilers would, in fact, not 
perform the addition at all in test_add.  Certainly we are not intending 
to imply that this common optimization is now illegal.  Subroutine 
test_add contains no references to the IEEE modules.  I would prefer to 
either delete this paragraph from the ANSWER, or to delete the end of 
the sentence following "useful".   Further, the calls to the IEEE 
routines in the caller are a distraction from the point of the question, 
and could be deleted as well.

Cheers,
Bill


-- 
Bill Long                                           longb at cray.com
Fortran Technical Support    &                 voice: 651-605-9024
Bioinformatics Software Development            fax:   651-605-9142
Cray Inc./Cray Plaza, Suite 210/380 Jackson St./St. Paul, MN 55101