This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: new test for type extensions from f2003 std.
- From: Richard E Maine <Richard dot Maine at nasa dot gov>
- To: Uttam Pawar <uttamp at us dot ibm dot com>
- Cc: fortran at gcc dot gnu dot org
- Date: Mon, 21 Nov 2005 11:18:54 -0800
- Subject: Re: new test for type extensions from f2003 std.
- References: <1132597105.32717.14.camel@linux.site>
On Nov 21, 2005, at 10:18 AM, Uttam Pawar wrote:
Here is a new test for type extensions (Reference f2003 std. 4.5.6.1
Inheritance). I believe it is not supported yet.
I'm a little confused. You expected anything else in an f95 compiler?
This is part of what is arguably the biggest and most complicated new
feature in f2003. While I'd love to see it supported, that's because
I'd love to see an f2003 compiler. By the time you have that stuff
done, you would be most of the way to having an f2003 compiler and
could, with some justification, start referring to it as an f2003
compiler that wasn't yet feature complete, rather than an f95 compiler
with a few f2003 features.
Sort of like an f77 compiler that added on f90 modules. There basically
never were any of those because modules were so big that by the time
you had them done, you might as well finish f90.
Type extension isn't quite that extreme because you actually can do
type extension without adding in the rest of the object-oriented stuff.
The latest release of the NAG compiler does that. But I also noticed,
in testing out the NAG compiler, that it doesn't actually do much good
to have type extension without some of the other object-oriented
features. I could define some extended types, but it turned out that
whenever I wanted to actually do anything useful with them, I quickly
ran into needing other of the f2003 features (that NAG hadn't yet
done).
--
Richard Maine | Good judgment comes from experience;
Richard.Maine@nasa.gov | experience comes from bad judgment.
| -- Mark Twain