Ondřej Čertík

No, on macOS it is immediate. And in Ubuntu it used to be immediate.

It's so slow it is unusable. Let's disable stacktraces on linux. We'll just use gdb or lldb there. Then let's enable the llvm based stacktraces which are fast. TODO: *...

@gxyd if you think it will be too much trouble, then let's do it later. The priority is to merge the simplifier pass.

Ok, let's do it later. @kmr-srbh I apologize for not making this decision sooner. We'll resurrect it once the simplifier branch is merged.

I don't know what happened here: I used the online "resolve conflicts" feature, and to create a new branch. And it instead retargeted this PR to be from the new...

Closing in favor of https://github.com/lfortran/lfortran/pull/5182.

Recent discussion at Discourse: https://fortran-lang.discourse.group/t/whats-your-experience-with-using-for-accessing-object-members/6642. The old discussion is at the original gitlab issue above, please read it in full before commenting here.

If LFortran supports `.`, it can also automatically convert it to `%`, as requested, e.g., here: https://fortran-lang.discourse.group/t/changing-to-to-access-components-of-derived-types/8726

I would say: let's first figure out how to make the CI pass with your approach. Then let's figure out how to still use compile time values, since we want...

> How do I add the lfortran_intrinsics_iso_fortran_env.f90 file to the runtime library that gets linked? This suggests you converted it to runtime, see https://github.com/lfortran/lfortran/pull/4868#discussion_r1769610885. We should not do runtime. However,...