tech-pkg archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Concensus on PKGSRC_FORTRAN?=gfortran?



On 2020-03-26 10:10, Greg Troxel wrote:
Jason Bacon <outpaddling%yahoo.com@localhost> writes:

On 2020-03-26 09:58, Greg Troxel wrote:
Jason Bacon <outpaddling%yahoo.com@localhost> writes:

I think the only thing to watch out for is people using platforms
where the gcc packages don't build, but some of their builds require
Fortran.  As long as g95 is still available, we'd only be forcing them
to override a default, though, not taking away their ability to
function.
And, we can leave OS/version/platform combinations at g95, when we know
that's better.

Is this a theoretical concern, or an actual known problem?
This was raised as a reason not to default to gfortran years ago, but
it's only theoretical AFAIK at this point.  gfortran works well on all
the platforms I use.
So how about:

   during the freeze we don't change anything

   people doing bulk builds are welcome to flip to gfortran (really,
   anybody doing builds is welcomm to do whatever they want anyway!)

   we see if we get any more feedback

   close after the freeze, we flip the default to gfortran, unconditionally

   as we get reports of trouble, we put in conditionals to put g95 back
   for platforms where that makes sense.  (perhaps we add a .mk for this
   to share between gcc and clang)
Sounds reasonable to me.



Home | Main Index | Thread Index | Old Index