Should we avoid assignment of derived types in robust programs?

[Beliavsky]

@ivanpribec

Yes, the verbosity of associate can be a nuisance. (But not necessarily worse than restricting yourself to calling subroutines everywhere.) Perhaps the standard could introduce a second form of associate, which implicitly terminated at the end of functions, subroutines, and blocks (I think there was a thread about this).

The thread was here, and there is an issue A shorthand for immutability at j3-fortran/fortran_proposals.