[Bug fortran/80666] character length parameter fails if declaration order incorrect

["dominiq at lps dot ens.fr" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80666

--- Comment #9 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
> Yes I would object to closing it.
> This issue has a large impact on significant bodies of legacy code
> in our institute.

Common! How many of such errors do you see now?

> I understand that you wish to enforce "standards",

I don't want to "enforce" anything, I am just trying to reflect the present
situation of gfortran: 908 PRs opened as for today (may be a dozen fixed
pending) and a shortage of man-power.

In my opinion, it would be fair with respect to the gfortran users to close as
WONTFIX PRS expected to rot forever.

> but removing a "Non-standard extension" that was present in older
> gfortran versions is wrong in my opinion.

IMO what is wrong is

(1) use extensions
(2) expect that they will work at time t+1 because they "worked" at time t.

> You should at least leave a commandline option or environment switch
> to allow users to compile legacy code that worked fine before.

IMO you have the following options:

(1) Fix your code.
(2) Stick to an old version that handle your pet extension.
(3) Do the work (or pay someone to do it) to restore the extension (expect some
fight to get it accepted).
(4) Wait forever a fix which will never come.