Wish List for Fortran after 2015 from Van Snyder

[certik]

Van has collected feedback over the past 50 years and with his permission, I am posting his document here:

• wishlist.pdf (November 17, 2017)
• wishlist.pdf (February 2, 2022)

I plan to extract all ideas from the document and create issues here. Many of the ideas are interesting, but would have to be developed more. Some of the ideas were rejected by the committee, but I would still like to create an issue for it, and then write up arguments why it was rejected, along with the link to the formal proposal that was submitted to the committee. So that future proposals can learn from the previous work.

1 Important

• 1.1“Physical” or “Engineering” Units of Measure (“Physical” or “Engineering” Units of Measure #50)
• 1.2 Improvements to the type system
  • 1.2.1 A more complete type system should be provided (Enumerations that are true types, and their enumerators #51, Subranges of integers #52)
  • 1.2.2 New types – not type synonyms – from existing types
• 1.3 Updaters
• 1.4 Uniform syntax (Uniform syntax #53)
• 1.5 Exception generation and handling
• 1.6 Coroutines
• 1.7 Generic programming
• 1.8 An assertion system would have several benefits
• 1.9 Structured parallelism would assist to exploit modern architectures

2 Useful

• 2.1 Improvements to constructs
  • 2.1.1 All constructs should be allowed to have specification parts
  • ...
• ...