RFC: future gfortran development and subversion

[Bernhard Fischer]

On Fri, Oct 21, 2005 at 11:06:07PM +0200, Paul Thomas wrote:
>Thomas,
>
>>Currently, gfortran is in a half-usable state.  It is not yet
>>ready as a replacement for g77 (see PR 19292) and there are quite
>>a lot of things it gets wrong with Fortran 95.
>>
>I think that this is way too strong.   Of the outstanding PR19292 "bugs":

>16465 parser chokes with ffixed-line-length-7 - is this essential?

21302 Max line length in free form mode

IIRC someone suggested to rename -ffixed-line-length- to -fline-length-
(with a backwards compatible alias for the former) and use the given
value for both forms.

If we consider that somebody may want to pass both free- and
fixed-form source to one compiler invocation, then having two separate
arguments for fixed and freeform line-lengths may be desirable.
e.g. default line-len for free-form but unlimited for fixed form would
use this arguments:
$ gfortran-4.1-HEAD -o foo f1.F f2.f90  -ffixed-line-length-none
cc1: error: argument to "-ffixed-line-length-" should be a non-negative
integer

Should the argument for maximal line length for freeform be distinct
from the argument for maximal line length in fixed form?


Should the default line-length for free form default to the full line?


e.g the example from pr21302:
$ gfortran-4.1-HEAD -o a.elf f1.F f2.f90 -ffixed-line-length-none
-Wline-truncation
cc1: warning: command line option "-Wline-truncation" is valid for
Fortran but not for C
cc1: error: argument to "-ffixed-line-length-" should be a non-negative
integer
 In file f2.f90:3

).gt.
    1
Error: Syntax error in expression at (1)
 In file f2.f90:3

).gt.
    1
Warning: Line truncated at (1)
 In file f2.f90:4

The paperwork is on it's way back to the FSF, so will hopefully be
finished soon, fwiw.
I can send a proposal as soon as i get the ack from the copyright
assignment.

>I would say that this is a 33:33:33 mix of fixed:do-able:ignorable/wrong - 
>let's have a blitz on it.