Fortran blog post with Institute for Research Software

Earlier this year, I tried to bring together some geoscience Fortranners at the EGU conference. It’s been my experience that many Fortran users in geoscience are quite oblivious of the developments of the past 5+ years – ecosystem improvements in particular.

It was a good chat. Jack took some notes and suggested writing something up [Thanks for the initiative, Jack]. We wrote a discussion summary that is now a blog post with the Institute for Research Software.

5 Likes

A mention is made about a lack of example programs for intrinsics. Still being expanded, but see
GitHub - urbanjost/M_intrinsics: man-page style descriptions of Fortran intrinsics for use as a reference for developers and tutorials · GitHub which has descriptions as a slidy HTML document, plain HTML, manpages, a CLI interface with optional color highlighting — the most common intrinsics all have working standalone example programs. The fman command lets you extract the demo programs but they are also available in the repository and include a make and fpm interface for building them. One of the documents will pass the demo programs into the playground page but that has not worked for quite a while. On the to-do list is to figure out how to pass the example programs to Compiler Explorer and the LFortran demo page as is done with the playground page.
They have not been updated for a while, but the descriptions also appear on Fortran.lang but need an update. Perhaps not a major piece of infrastructure but since a lack of intrinsic demos was mentioned I thought I would point out that is available. I am surprised there was no mention of the lack of free documentation compared to many other languages. Fortran beginners where I work are always surprised on how much less documentation is available on-line as there is for many other languages. I hate to point out it has improved dramatically compared to previous decades.

1 Like

Thanks for linking the documentation project, @urbanjost. This hadn’t been on my radar, I must admit, which in turn underlines the point about the Fortran community doing great work that remains somewhat invisible to others. (I even follow you on GitHub).

You’re very right about the lack of free documentation. It was touched on or circled around somewhat with the comments about Stack Overflow and documentation of intrinsics, but agree it deserves more attention. I also notice newcomers being surprised about the relative lack of documentation and tutorials. It has definitely gotten better, but we’re not there yet. It’s one reason I’m currently translating the tutorials of a free online course I created – an intro to climate science and short coding crash course for complete newcomers – into Fortran. (I do this while using it in teaching and getting feedback from students, who chose to use Fortran instead of Python).