Is there a way to register radian definitions correctly?

[Abdul-Lateef Haji-Ali]

I like that Radian compiles my config file as part of radian.el (in fact I made a change to Radian's code to compile all files in a custom init directory as part of radian.el).

However, one problem is that all functions/variables are registered to have their definition in radian.el rather than in the correct file where they were defined. I know that the compilation process of radian embeds my init files as part of radian.el, but I am wondering if there is some emacs-wizardy that we can do to correct for this issue.

Thoughts?

[Radon Rosborough]

Huh, you're right. I had never even noticed.

[Abdul-Lateef Haji-Ali]

That's much better thanks!

I am still seeing package functions like no-littering-expand-var-file-name being shown to be defined in radian.el (for example if I call helpful-function on it.
Is it just me?

I am not 100% what the cause is, except that no-littering-expand-var-file-name appears twice in load-history.

This is the value of load-history:

(...

(".../radian/emacs/radian.elc"
   ...
   (defun . no-littering-expand-etc-file-name)
   (defun . no-littering-expand-var-file-name)
   (provide . no-littering-autoloads)
  ...)
...
(".../straight/build-29.0.90/no-littering/no-littering.elc"
  (require . cl-lib)
  (require . compat)
  no-littering-etc-directory no-littering-var-directory
  (defun . no-littering-expand-etc-file-name)
  (defun . no-littering-expand-var-file-name)
  (provide . no-littering))
...) 

Maybe there's an issue with autoloads being considered the true definition? Not sure this is radian related or use-package related.

[Abdul-Lateef Haji-Ali]

Sorry for the noise. Radian is behaving as expected.
Emacs needs a better way of finding the source of a function that is being autoloaded.