GCC 4.6.0 released

G95 was not included in GCC by choice of its developer, and that was the reason for the G95/gfortran fork in the first place, or there wouldn't have been two GCC-based Fortran compilers.

G95 was never the "preferred front end" because it was never part of GCC. The latest stable G95 is still based on GCC 4.0, but the project appears to be abandoned. On the other hand, the folks working on gfortran continue to do an amazing job, especially considering their small numbers and the fact that they are all volunteers (mostly engineers and physicist).

Got a link to that NASA chief's complaint? There are NASA folks and many others on comp.lang.fortran who are talking generally favorable about gfortran, and *especially* about the quick replies to bug reports. And two US national laboratories (LLNL and LANL) have made significant contributions to gfortran over the years.

For Fortran 2003/2008 support: patches welcome. Note that there is only one compiler (Cray) that fully implements Fortran 2003, and not even a single one that fully implements Fortran 2008. Some others claim to support Fortran 2003 at least, but last I checked, there is always still one thing missing from the complete language. This has nothing to do with politics and bickering. It is just a lot of work because the changes from Fortran 95 to Fortran 2003 are huge, comparable to the changes from C to early versions of C++. That is apparently a challenge even for the commercial vendors.