Re: g77/Mountain Lion/XCode 4.6

[Janus Weil <janus at gcc dot gnu dot org>]

Hi Zack,

> I've done a fair amount a research, but most solutions are to "stop using g77". And while I agree that that would be the ideal solution, the PI and grad students are adamant they need g77 to compile some of their older code.

at this point in time, "stop using g77" is pretty much the best advice
you can get. Seriously.

If your code compiles with g77, but not with gfortran, chances are
high that it either uses non-standard Fortran or a very old dialect.

In both cases, my first advice would be to update your code: Remove
non-standard features and make it compile with other compilers
(gfortran, ifort, sunf95, g95, ...).

If that fails for some reason, or your code is so large and messy that
you don't see a chance of fixing it, or you are just too lazy, your
best bet is to file a bug report for gfortran and request your
non-standard feature to be supported as an extension (e.g. with
-std=legacy). That may work or may not, depending on whether you can
convince one of the gfortran developers (who are basically all
volunteers, working on the compiler in their spare time) to implement
it.

If this also fails, you (or one of your grad students) can still
invest some work to implement the missing feature yourself.


> Any help would be very much appreciated. If there is another/better avenue to request support, I'd be happy to follow those recommendations as well. I can't be the only person trying to get g77 to compile on a freshly installed copy of mountain lion, right?

I seriously hope you are!

Btw, this is the gfortran mailing list. Which means you can expect to
get large amounts of help with anything that concerns gfortran, but
I'm rather sceptical  whether anyone here is willing to invest any
time in helping you make (the really really outdated) g77 run on your
system.

IMHO the best way to proceed would be to tell us exactly why your
codes don't compile with gfortran.

Cheers,
Janus