Do you think resorting to a preprocessor could resolve my problem and define a constant literal linked to an array of strings with items of different length?
10.1.5.1 Intrinsic operation classification, para 6, line 27-28:
The kind type parameters of the operands of a character relational intrinsic operation shall be the same.
Line 9 is always wrong. Depending on whether c1 or c2 corresponds to default kind, Line 10 or 11 will be wrong.
I stand corrected. Thanks for the info. I had just assumed that because 1.0e0 == 1.0d0 was valid, something similar would be for characters as well. But once again we all see what happens when you assume…
1 Like
Welcome to the club (of the unfortunate 
)
See my comments in this thread, especially " How easy and convenient is to develop libraries in Fortran that can used effectively and productively by the broader scientific and technical community? In year 2020, the reality is it is extremely difficult, impossible even in some situations. This extends to both standard libraries as well as other community / domain-specific libraries. The issues here are another indicators of where the language needs to improve."
Add the discussion here with what should be trivially simple string_type development in a modern language to the issues I mentioned in the thread above.
We, in industry, realized the limitations mid to late 2000s given so many loose-ends in the 2003 edition.
Though Fortran 2008 gave some hope, the next revision 2018 nearly a decade later and the plan toward 202X has proved to be so disappointing to many powers-that-be that so much of technical and scientific computing has migrated away.
As I have commented on this forum as well as others (comp.lang.fortran, GitHub j3-fortran site, Intel Forum, etc.) I am of the firm view the limitations with the language are entirely borne out of aspects missing or the baggage persisting in the standard.
3 Likes