RFC: future gfortran development and subversion

[Paul Thomas]

FX,

>
> The fortran patches are always fortran-contained, and I think if the 
> community thinks it worth to have a different development model (until 
> some point in the future, defined in advance) why shouldn't it be so?
>
This might well be the value of keeping the binaries going.  From what I 
can see, most people using gfortran in anger are updating from your 
site, rather than using the distros.

I think that keeping two lines going is a pain but just about OK; better 
would be to undertake the periodic synchronizations from Head to 4.1, 
even if we are a bit resource limited.  As it is, this latter makes for 
some of the divergence because not all patches make their way to the 
distribution version.

By the way, I just posted on the Wiki a comparison of DF6.0 and gfortran 
using the Polyhedron benchmarks.  It's not too bad at all but there are 
a couple of sore thumbs.  I have just done the ifc to gfortran 
comparison on FC3, which basically looks the same (one bloody thumb is 
repaired because the comparison is not between a native user of system 
resources and another that does so via Cygwin.).  I will try to collate 
and post the results in a day or so.

Paul