Improving Fortran standardization process (lessons from C++23 getting multidimensional arrays)

42

At this stage given all that has transpired, I think it will be infinitely better to focus on addressing the gaps and limitations and improving the areas of weaknesses.

As part of this, I would urge everyone practicing Fortran in their computational efforts to contemplate For whom Fortran, for what.

I personally think Fortran must strive to remain a general-purpose programming paradigm with particular focus on scientific and technical computing covering HPC (exascale and beyond) to smaller scale in research, industry, and academic from hobbyist/single developer to very large teams across platforms. A standard really helps with such broad needs and Fortran already has one.

What Fortran lacks immediately is resources: contributors and funds.

With more resources of the right kind, considerable improvements can be made. It does not always solve the problem but resources are a must, there can be little hope otherwise.

2 Likes
43

QED.

If anyone is confused by the brevity of the post (as is the discourse bot), QED means I told you so.

44

@PetrKryslUCSD and others, we are doing exactly that, with fortran-lang projects here (fpm, stdlib, discussion forum, community, etc.) as well as a central repository for ideas for Fortran improvement: GitHub - j3-fortran/fortran_proposals: Proposals for the Fortran Standard Committee and we are also working towards a community compiler or compilers.

What exactly has been stopping any compiler developer, commercial or open source, from doing that for the last 30 years?

Nothing, except that it is a lot of work. But we are doing it, and I welcome anyone here to help us get over the finish line sooner.

If you want to help with LFortran, you can do it here: GitHub - lfortran/lfortran: Official main repository for LFortran, we are making excellent progress and can always use more hands.

3 Likes